Menu
▾
▴
astlinux-devel — The AstLinux development environment, compilation, etc
You can subscribe to this list here.
| 2008 |
Jan
|
Feb
|
Mar
(41) |
Apr
(35) |
May
(18) |
Jun
(5) |
Jul
(4) |
Aug
(37) |
Sep
(9) |
Oct
(20) |
Nov
(50) |
Dec
(217) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2009 |
Jan
(212) |
Feb
(76) |
Mar
(113) |
Apr
(88) |
May
(130) |
Jun
(54) |
Jul
(208) |
Aug
(223) |
Sep
(112) |
Oct
(63) |
Nov
(131) |
Dec
(103) |
| 2010 |
Jan
(247) |
Feb
(130) |
Mar
(43) |
Apr
(92) |
May
(40) |
Jun
(43) |
Jul
(43) |
Aug
(80) |
Sep
(44) |
Oct
(74) |
Nov
(21) |
Dec
(46) |
| 2011 |
Jan
(36) |
Feb
(11) |
Mar
(21) |
Apr
(33) |
May
(4) |
Jun
(12) |
Jul
(5) |
Aug
(20) |
Sep
|
Oct
(64) |
Nov
(26) |
Dec
(71) |
| 2012 |
Jan
(13) |
Feb
(24) |
Mar
(11) |
Apr
(2) |
May
(10) |
Jun
(5) |
Jul
(13) |
Aug
(7) |
Sep
(26) |
Oct
(22) |
Nov
(17) |
Dec
(16) |
| 2013 |
Jan
(6) |
Feb
(6) |
Mar
(6) |
Apr
(8) |
May
(20) |
Jun
|
Jul
(1) |
Aug
(4) |
Sep
(18) |
Oct
(3) |
Nov
(14) |
Dec
(33) |
| 2014 |
Jan
(26) |
Feb
(6) |
Mar
(69) |
Apr
(10) |
May
|
Jun
(8) |
Jul
(18) |
Aug
(22) |
Sep
(19) |
Oct
(17) |
Nov
|
Dec
(4) |
| 2015 |
Jan
(14) |
Feb
(18) |
Mar
|
Apr
|
May
(26) |
Jun
(8) |
Jul
(9) |
Aug
(10) |
Sep
(15) |
Oct
(2) |
Nov
(30) |
Dec
(33) |
| 2016 |
Jan
(1) |
Feb
(24) |
Mar
(19) |
Apr
(1) |
May
|
Jun
(3) |
Jul
(1) |
Aug
(1) |
Sep
(20) |
Oct
(5) |
Nov
(14) |
Dec
(4) |
| 2017 |
Jan
(15) |
Feb
(35) |
Mar
(10) |
Apr
(9) |
May
(14) |
Jun
(33) |
Jul
(1) |
Aug
(27) |
Sep
(7) |
Oct
|
Nov
(10) |
Dec
(15) |
| 2018 |
Jan
(29) |
Feb
|
Mar
(2) |
Apr
(1) |
May
(11) |
Jun
|
Jul
(1) |
Aug
(8) |
Sep
(11) |
Oct
(22) |
Nov
(9) |
Dec
(13) |
| 2019 |
Jan
(1) |
Feb
(7) |
Mar
(3) |
Apr
(21) |
May
(34) |
Jun
(36) |
Jul
(18) |
Aug
(17) |
Sep
(19) |
Oct
(8) |
Nov
(3) |
Dec
|
| 2020 |
Jan
|
Feb
(4) |
Mar
(8) |
Apr
(29) |
May
(50) |
Jun
(8) |
Jul
(2) |
Aug
(10) |
Sep
(1) |
Oct
(7) |
Nov
(9) |
Dec
(19) |
| 2021 |
Jan
(2) |
Feb
(9) |
Mar
(6) |
Apr
(21) |
May
(13) |
Jun
(11) |
Jul
(2) |
Aug
(1) |
Sep
(3) |
Oct
(26) |
Nov
(2) |
Dec
(16) |
| 2022 |
Jan
(8) |
Feb
(7) |
Mar
(1) |
Apr
(13) |
May
(1) |
Jun
(4) |
Jul
(4) |
Aug
(1) |
Sep
(1) |
Oct
|
Nov
|
Dec
(1) |
| 2023 |
Jan
(2) |
Feb
(3) |
Mar
(16) |
Apr
|
May
(2) |
Jun
(1) |
Jul
(4) |
Aug
(13) |
Sep
(8) |
Oct
(6) |
Nov
(4) |
Dec
|
| 2024 |
Jan
(3) |
Feb
(3) |
Mar
|
Apr
|
May
(1) |
Jun
|
Jul
(5) |
Aug
|
Sep
(1) |
Oct
|
Nov
(5) |
Dec
|
| 2025 |
Jan
(4) |
Feb
(2) |
Mar
|
Apr
(11) |
May
(1) |
Jun
(9) |
Jul
(18) |
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
|
From: Ramesh GK <ram...@ho...> - 2020-12-28 11:31:32
|
Sure, Lonnie. Will do as suggested and keep posted. Also i read through online that i586 motherboards are no longer manufactured/supported and x86_64 is the only architecture supported mostly going forward. Thanks Ramesh GK Phone: 408-462-0408 Strive not to be a success but rather to be of value - Albert Einstein ________________________________ From: Lonnie Abelbeck <li...@lo...> Sent: Sunday, December 27, 2020 7:25 PM To: AstLinux Developers Mailing List <ast...@li...> Subject: Re: [Astlinux-devel] Linux 2.6.x Support Hi Ramesh, > and also thinking to call the builds as "AstLinux 1.3.8.yyyymmdd (Home Edition)" unless there are any concerns. I think keeping "AstLinux" in your project title could be very confusing since your focus is quite different. Though, a reference in your documentation or repository page showing your project is a "fork of" or "based upon" AstLinux would be appreciated. Lonnie > On Dec 27, 2020, at 6:04 PM, Ramesh GK <ram...@ho...> wrote: > > Hi Lonnie, > > Thanks for the feedback. I was trying to use kernel 2.6.x and got a fair amount of lessons learnt which helped me with figuring out things not only from a kernel perspective but also with the limitations of gcc 4.4 compiler (just for my lessons) which made me to downgrade some of the packages. I was in fact successful in building all the software however had hit with a bottleneck of serial devices hardware support and mandatory compiler upgrade needed to support for the additional software I am using. > > From a direction perspective, I am planning to go with a thumb rule of keeping back the astlinux-1.3.8 packages as-is (mostly) to the original including the kernel (updating to 3.16.85) with some changes to the toolchain and any fixes to the build that is needed. One of the reasons to stick with 1.3.8 version is cross platform support offered for i586 platforms and to use the combined partition model to support for the additional software I am adding. > > To give some clarification, I am adding packages that are typically needed for a Home Environment but not limited to other environments. > For example - proxy (squid, privoxy, e2gaurdian/dansguardian), web server (lighttpd/nginx), mail server (exim, dovecot), home security/automation software (motion, ser2net, openhab), wireless capabilities (hostapd, qmi, mbim) etc > > In addition, I am also planning to keep it in sync with the astlinux road map for software packages as time permits and also thinking to call the builds as "AstLinux 1.3.8.yyyymmdd (Home Edition)" unless there are any concerns. > > Thanks > Ramesh GK > Phone: 408-462-0408 > > Strive not to be a success but rather to be of value - Albert Einstein > > From: Lonnie Abelbeck <li...@lo...> > Sent: Friday, December 25, 2020 8:33 PM > To: AstLinux Developers Mailing List <ast...@li...> > Subject: Re: [Astlinux-devel] Linux 2.6.x Support > > Hi Ramesh, > > Quick answer, the most recent runnix versions should boot most any AstLinux version, even a kernel 2.6.x version. Runnix uses kexec to boot an AstLinux run image. Yes a i368 runnix can boot a x86_84 AstLinux run image. > > Better answer, I am very perplexed why you want to go way back to kernel 2.6.x, it contains many security vulnerabilities. > > Also the cross-toolchain is built based on the kernel headers, currently 4.19.x . I'm not sure what issues you would have using the latest toolchain with a 2.6.x kernel ... the issues may be subtile. Some package builds may fail if they are tied to the kernel version. > > Also a issue, the unionfs in-kernel patch (2.6.x) vs. the current FUSE unionfs. > > I would suggest against using a 2.6.x kernel. > > Lonnie > > > > > On Dec 25, 2020, at 5:28 PM, Ramesh GK <ram...@ho...> wrote: > > > > Hi, > > > > I am trying to downgrade the kernel to 2.6.x for squeezing the performance out of the system and currently working on AstLinux 1.3.8. Most of the packages are looking good except for the drivers which I have downgraded. > > > > Most of the work looks good except for runnix which looks to be a binary. I am not sure if runnix 0.5.11 will work kernel 2.6.35.14 and tried to get runnix 0.4-6527 however it looks to i386 only. > > > > Would it be feasible to provide a runnix 0.5.11 or 0.4-6527 with x86_64 for kernel 2.6.35.14? > > > > (or) if you can let me know how to generate that package then I can try it as well. > > > > Appreciate your help in advance. > > > > Thanks > > Ramesh GK > > Phone: 408-462-0408 > > > > Strive not to be a success but rather to be of value - Albert Einstein > > _______________________________________________ > > Astlinux-devel mailing list > > Ast...@li... > > https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.sourceforge.net%2Flists%2Flistinfo%2Fastlinux-devel&data=04%7C01%7C%7Cd7ef5b55fa9a44a51e5308d8aac71b3e%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637447119413574448%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=4eVRgbk9w9sJzJzMngqi1KLL3pugHO%2F7iQUxqKpQO%2FY%3D&reserved=0 > > > > _______________________________________________ > Astlinux-devel mailing list > Ast...@li... > https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.sourceforge.net%2Flists%2Flistinfo%2Fastlinux-devel&data=04%7C01%7C%7Cd7ef5b55fa9a44a51e5308d8aac71b3e%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637447119413584442%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=ev9EFLlZQ0zEe1k3VyCpAY%2FF6C4%2BKfutaVAmbjMmFVU%3D&reserved=0 > _______________________________________________ > Astlinux-devel mailing list > Ast...@li... > https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.sourceforge.net%2Flists%2Flistinfo%2Fastlinux-devel&data=04%7C01%7C%7Cd7ef5b55fa9a44a51e5308d8aac71b3e%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637447119413584442%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=ev9EFLlZQ0zEe1k3VyCpAY%2FF6C4%2BKfutaVAmbjMmFVU%3D&reserved=0 _______________________________________________ Astlinux-devel mailing list Ast...@li... https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.sourceforge.net%2Flists%2Flistinfo%2Fastlinux-devel&data=04%7C01%7C%7Cd7ef5b55fa9a44a51e5308d8aac71b3e%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637447119413584442%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=ev9EFLlZQ0zEe1k3VyCpAY%2FF6C4%2BKfutaVAmbjMmFVU%3D&reserved=0 |
|
From: Lonnie A. <li...@lo...> - 2020-12-28 00:25:28
|
Hi Ramesh, > and also thinking to call the builds as "AstLinux 1.3.8.yyyymmdd (Home Edition)" unless there are any concerns. I think keeping "AstLinux" in your project title could be very confusing since your focus is quite different. Though, a reference in your documentation or repository page showing your project is a "fork of" or "based upon" AstLinux would be appreciated. Lonnie > On Dec 27, 2020, at 6:04 PM, Ramesh GK <ram...@ho...> wrote: > > Hi Lonnie, > > Thanks for the feedback. I was trying to use kernel 2.6.x and got a fair amount of lessons learnt which helped me with figuring out things not only from a kernel perspective but also with the limitations of gcc 4.4 compiler (just for my lessons) which made me to downgrade some of the packages. I was in fact successful in building all the software however had hit with a bottleneck of serial devices hardware support and mandatory compiler upgrade needed to support for the additional software I am using. > > From a direction perspective, I am planning to go with a thumb rule of keeping back the astlinux-1.3.8 packages as-is (mostly) to the original including the kernel (updating to 3.16.85) with some changes to the toolchain and any fixes to the build that is needed. One of the reasons to stick with 1.3.8 version is cross platform support offered for i586 platforms and to use the combined partition model to support for the additional software I am adding. > > To give some clarification, I am adding packages that are typically needed for a Home Environment but not limited to other environments. > For example - proxy (squid, privoxy, e2gaurdian/dansguardian), web server (lighttpd/nginx), mail server (exim, dovecot), home security/automation software (motion, ser2net, openhab), wireless capabilities (hostapd, qmi, mbim) etc > > In addition, I am also planning to keep it in sync with the astlinux road map for software packages as time permits and also thinking to call the builds as "AstLinux 1.3.8.yyyymmdd (Home Edition)" unless there are any concerns. > > Thanks > Ramesh GK > Phone: 408-462-0408 > > Strive not to be a success but rather to be of value - Albert Einstein > > From: Lonnie Abelbeck <li...@lo...> > Sent: Friday, December 25, 2020 8:33 PM > To: AstLinux Developers Mailing List <ast...@li...> > Subject: Re: [Astlinux-devel] Linux 2.6.x Support > > Hi Ramesh, > > Quick answer, the most recent runnix versions should boot most any AstLinux version, even a kernel 2.6.x version. Runnix uses kexec to boot an AstLinux run image. Yes a i368 runnix can boot a x86_84 AstLinux run image. > > Better answer, I am very perplexed why you want to go way back to kernel 2.6.x, it contains many security vulnerabilities. > > Also the cross-toolchain is built based on the kernel headers, currently 4.19.x . I'm not sure what issues you would have using the latest toolchain with a 2.6.x kernel ... the issues may be subtile. Some package builds may fail if they are tied to the kernel version. > > Also a issue, the unionfs in-kernel patch (2.6.x) vs. the current FUSE unionfs. > > I would suggest against using a 2.6.x kernel. > > Lonnie > > > > > On Dec 25, 2020, at 5:28 PM, Ramesh GK <ram...@ho...> wrote: > > > > Hi, > > > > I am trying to downgrade the kernel to 2.6.x for squeezing the performance out of the system and currently working on AstLinux 1.3.8. Most of the packages are looking good except for the drivers which I have downgraded. > > > > Most of the work looks good except for runnix which looks to be a binary. I am not sure if runnix 0.5.11 will work kernel 2.6.35.14 and tried to get runnix 0.4-6527 however it looks to i386 only. > > > > Would it be feasible to provide a runnix 0.5.11 or 0.4-6527 with x86_64 for kernel 2.6.35.14? > > > > (or) if you can let me know how to generate that package then I can try it as well. > > > > Appreciate your help in advance. > > > > Thanks > > Ramesh GK > > Phone: 408-462-0408 > > > > Strive not to be a success but rather to be of value - Albert Einstein > > _______________________________________________ > > Astlinux-devel mailing list > > Ast...@li... > > https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.sourceforge.net%2Flists%2Flistinfo%2Fastlinux-devel&data=04%7C01%7C%7C226779e50af44f12aa3a08d8a93e6419%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637445432718557618%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=BvRyh%2BVKfs9%2BvJtu3eTpa2ISqoGqbzbLqStryJEGBG4%3D&reserved=0 > > > > _______________________________________________ > Astlinux-devel mailing list > Ast...@li... > https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.sourceforge.net%2Flists%2Flistinfo%2Fastlinux-devel&data=04%7C01%7C%7C226779e50af44f12aa3a08d8a93e6419%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637445432718567618%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=4yn%2BASFDLvjBoNL6Mcju00Dxj%2BfyKbZobJfY2zrETL4%3D&reserved=0 > _______________________________________________ > Astlinux-devel mailing list > Ast...@li... > https://lists.sourceforge.net/lists/listinfo/astlinux-devel |
|
From: Ramesh GK <ram...@ho...> - 2020-12-28 00:05:08
|
Hi Lonnie, Thanks for the feedback. I was trying to use kernel 2.6.x and got a fair amount of lessons learnt which helped me with figuring out things not only from a kernel perspective but also with the limitations of gcc 4.4 compiler (just for my lessons) which made me to downgrade some of the packages. I was in fact successful in building all the software however had hit with a bottleneck of serial devices hardware support and mandatory compiler upgrade needed to support for the additional software I am using. >From a direction perspective, I am planning to go with a thumb rule of keeping back the astlinux-1.3.8 packages as-is (mostly) to the original including the kernel (updating to 3.16.85) with some changes to the toolchain and any fixes to the build that is needed. One of the reasons to stick with 1.3.8 version is cross platform support offered for i586 platforms and to use the combined partition model to support for the additional software I am adding. To give some clarification, I am adding packages that are typically needed for a Home Environment but not limited to other environments. For example - proxy (squid, privoxy, e2gaurdian/dansguardian), web server (lighttpd/nginx), mail server (exim, dovecot), home security/automation software (motion, ser2net, openhab), wireless capabilities (hostapd, qmi, mbim) etc In addition, I am also planning to keep it in sync with the astlinux road map for software packages as time permits and also thinking to call the builds as "AstLinux 1.3.8.yyyymmdd (Home Edition)" unless there are any concerns. Thanks Ramesh GK Phone: 408-462-0408 Strive not to be a success but rather to be of value - Albert Einstein ________________________________ From: Lonnie Abelbeck <li...@lo...> Sent: Friday, December 25, 2020 8:33 PM To: AstLinux Developers Mailing List <ast...@li...> Subject: Re: [Astlinux-devel] Linux 2.6.x Support Hi Ramesh, Quick answer, the most recent runnix versions should boot most any AstLinux version, even a kernel 2.6.x version. Runnix uses kexec to boot an AstLinux run image. Yes a i368 runnix can boot a x86_84 AstLinux run image. Better answer, I am very perplexed why you want to go way back to kernel 2.6.x, it contains many security vulnerabilities. Also the cross-toolchain is built based on the kernel headers, currently 4.19.x . I'm not sure what issues you would have using the latest toolchain with a 2.6.x kernel ... the issues may be subtile. Some package builds may fail if they are tied to the kernel version. Also a issue, the unionfs in-kernel patch (2.6.x) vs. the current FUSE unionfs. I would suggest against using a 2.6.x kernel. Lonnie > On Dec 25, 2020, at 5:28 PM, Ramesh GK <ram...@ho...> wrote: > > Hi, > > I am trying to downgrade the kernel to 2.6.x for squeezing the performance out of the system and currently working on AstLinux 1.3.8. Most of the packages are looking good except for the drivers which I have downgraded. > > Most of the work looks good except for runnix which looks to be a binary. I am not sure if runnix 0.5.11 will work kernel 2.6.35.14 and tried to get runnix 0.4-6527 however it looks to i386 only. > > Would it be feasible to provide a runnix 0.5.11 or 0.4-6527 with x86_64 for kernel 2.6.35.14? > > (or) if you can let me know how to generate that package then I can try it as well. > > Appreciate your help in advance. > > Thanks > Ramesh GK > Phone: 408-462-0408 > > Strive not to be a success but rather to be of value - Albert Einstein > _______________________________________________ > Astlinux-devel mailing list > Ast...@li... > https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.sourceforge.net%2Flists%2Flistinfo%2Fastlinux-devel&data=04%7C01%7C%7C226779e50af44f12aa3a08d8a93e6419%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637445432718557618%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=BvRyh%2BVKfs9%2BvJtu3eTpa2ISqoGqbzbLqStryJEGBG4%3D&reserved=0 _______________________________________________ Astlinux-devel mailing list Ast...@li... https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.sourceforge.net%2Flists%2Flistinfo%2Fastlinux-devel&data=04%7C01%7C%7C226779e50af44f12aa3a08d8a93e6419%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637445432718567618%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=4yn%2BASFDLvjBoNL6Mcju00Dxj%2BfyKbZobJfY2zrETL4%3D&reserved=0 |
|
From: Lonnie A. <li...@lo...> - 2020-12-26 01:34:19
|
Hi Ramesh, Quick answer, the most recent runnix versions should boot most any AstLinux version, even a kernel 2.6.x version. Runnix uses kexec to boot an AstLinux run image. Yes a i368 runnix can boot a x86_84 AstLinux run image. Better answer, I am very perplexed why you want to go way back to kernel 2.6.x, it contains many security vulnerabilities. Also the cross-toolchain is built based on the kernel headers, currently 4.19.x . I'm not sure what issues you would have using the latest toolchain with a 2.6.x kernel ... the issues may be subtile. Some package builds may fail if they are tied to the kernel version. Also a issue, the unionfs in-kernel patch (2.6.x) vs. the current FUSE unionfs. I would suggest against using a 2.6.x kernel. Lonnie > On Dec 25, 2020, at 5:28 PM, Ramesh GK <ram...@ho...> wrote: > > Hi, > > I am trying to downgrade the kernel to 2.6.x for squeezing the performance out of the system and currently working on AstLinux 1.3.8. Most of the packages are looking good except for the drivers which I have downgraded. > > Most of the work looks good except for runnix which looks to be a binary. I am not sure if runnix 0.5.11 will work kernel 2.6.35.14 and tried to get runnix 0.4-6527 however it looks to i386 only. > > Would it be feasible to provide a runnix 0.5.11 or 0.4-6527 with x86_64 for kernel 2.6.35.14? > > (or) if you can let me know how to generate that package then I can try it as well. > > Appreciate your help in advance. > > Thanks > Ramesh GK > Phone: 408-462-0408 > > Strive not to be a success but rather to be of value - Albert Einstein > _______________________________________________ > Astlinux-devel mailing list > Ast...@li... > https://lists.sourceforge.net/lists/listinfo/astlinux-devel |
|
From: Ramesh GK <ram...@ho...> - 2020-12-25 23:29:15
|
Hi, I am trying to downgrade the kernel to 2.6.x for squeezing the performance out of the system and currently working on AstLinux 1.3.8. Most of the packages are looking good except for the drivers which I have downgraded. Most of the work looks good except for runnix which looks to be a binary. I am not sure if runnix 0.5.11 will work kernel 2.6.35.14 and tried to get runnix 0.4-6527 however it looks to i386 only. Would it be feasible to provide a runnix 0.5.11 or 0.4-6527 with x86_64 for kernel 2.6.35.14? (or) if you can let me know how to generate that package then I can try it as well. Appreciate your help in advance. Thanks Ramesh GK Phone: 408-462-0408 Strive not to be a success but rather to be of value - Albert Einstein |
|
From: David K. <da...@ke...> - 2020-12-22 03:04:40
|
The custom firewall rules that I provided earlier in this thread has a flaw
in that they require DNS to be active to resolve the FQDN to IP addresses.
This works if you restart the firewall after initial boot, but on boot up
the firewall is started before DNS services are available and so it fails.
Thanks to a tip from Lonnie I have updated my custom rule script...
open_for_host_port()
{
local host="$1"
local port="$2"
local type="$3"
echo "[CUSTOM RULE] spawn background to open port $port/$type for $host"
(
RETRY=6
while [ $RETRY -gt 0 ]; do
RETRY=$((RETRY - 1))
local IFS=$'\n'
local IP=""
local IPV4="$(host -t A $host | sed -n -r -e 's#^.* has address
([0-9.]+)$#\1#p')"
if [ -n "$IPV4" ]; then
for IP in $IPV4; do
IP="${IP##* }"
echo "[CUSTOM RULE] open port $port/$type for $host at $IP"
ip4tables -I EXT_INPUT_CHAIN -s $IP -p $type -m $type --dport
$port -j ACCEPT
done
local IPV6="$(host -t AAAA $host)"
for IP in $IPV6; do
IP="${IP##* }"
echo "[CUSTOM RULE] open port $port/$type for $host at $IP"
ip6tables -I EXT_INPUT_CHAIN -s $IP -p $type -m $type --dport
$port -j ACCEPT
done
exit
else
echo "[CUSTOM RULE] DNS service not up, sleep for 5 seconds"
logger -t CUSTOM_RULE -p user.info "DNS service not up, sleep for 5
seconds"
sleep 5
fi
done
echo "[CUSTOM RULE] Time out waiting for DNS service, ports not opened
for $host"
logger -t CUSTOM_RULE -p user.error "Time out waiting for DNS service,
ports not opened for $host"
) &
}
open_for_host_port "all.sipis.acrobits.cz" "5061" "tcp"
On Sat, Nov 28, 2020 at 6:42 PM Michael Knill <
mic...@ip...> wrote:
> Thanks David
>
>
>
> Yes I have used Bria with Wireguard VPN and it works fine. Also fails over
> well when roaming between Wifi and 4G. Yes you need to manually activate it
> which is annoying and I think it may prevent the softphone from
> backgrounding so battery problems as well.
>
>
>
> Interestingly however, I did some testing this morning with my Acrobits
> portal and the Cloud Softphone product on a test account using TLS.
> Provisioning was a matter of pointing the phone to a QR code and it worked
> first time.
>
> There appeared to be no issues with re-registration between the push
> server and direct and I did a test driving down the road and there was a
> short break but it roamed successfully between Wifi and 4G. Bit annoying as
> the call only stays up for 1m so not really a long enough test but long
> enough to notice successful roaming.
>
>
>
> I think its time to test it for real now. I will keep you posted. Thanks
> all for the feedback.
>
>
>
> Regards
>
> Michael Knill
>
>
>
> *From: *David Kerr <da...@ke...>
> *Reply to: *AstLinux Developers Mailing List <
> ast...@li...>
> *Date: *Sunday, 29 November 2020 at 9:51 am
> *To: *AstLinux Developers Mailing List <
> ast...@li...>
> *Subject: *Re: [Astlinux-devel] Recommendations for my architecture
> moving forward
>
>
>
> Looking at
> https://developer.apple.com/documentation/networkextension/personal_vpn/vpn_on_demand_rules
> it seems that Wireguard have implemented very basic VPN on-demand (turn it
> on when on cellular or when on wifi). What would be needed is much more
> granular control, at a minimum turn on for specific applications only.
> Even better would be connection specific rules. That feels like a lot of
> work which would have to get implemented in the Wireguard iOS app.
>
>
>
> David
>
>
>
> On Sat, Nov 28, 2020 at 5:35 PM David Kerr <da...@ke...> wrote:
>
> I used to have 5061 open to anything, but I started to get attempts to
> login and make calls, so I now restrict it to the Acrobits servers. For
> the iPhone itself I have to be connected to my home network either directly
> to the WiFi or by VPN (I use Wireguard). Acrobits (afaik) has no built-in
> VPN. I have not tried the Wireguard on-demand setting but it is possible
> that might work.
>
>
>
> I have not tried roaming between WiFi and Mobile.
>
>
>
> David
>
>
>
> On Sat, Nov 28, 2020 at 3:44 PM Michael Knill <
> mic...@ip...> wrote:
>
> Hi All
>
>
>
> Thanks for responding.
>
> Bria is certainly the gold standard however I have been down the Acrobits
> path and had a full demo of what they can provide. They have a full white
> labelling solution and there provisioning system is pretty cool. Just point
> your phone to a QR code.
>
>
>
> David are you only opening up 5061 to the push server? If so then your
> iPhone must never directly register to your system externally?
>
> Also another question; can you roam between Wi-Fi and 4G while on a call?
>
>
>
> Regards
>
> Michael Knill
>
>
>
> *From: *David Kerr <da...@ke...>
> *Reply to: *AstLinux Developers Mailing List <
> ast...@li...>
> *Date: *Sunday, 29 November 2020 at 2:48 am
> *To: *AstLinux Developers Mailing List <
> ast...@li...>
> *Subject: *Re: [Astlinux-devel] Recommendations for my architecture
> moving forward
>
>
>
> I have had Acrobits Softphone installed on my iPhone for several years. I
> don't use it a whole lot but it works well. It has push notification for
> when the softphone is "in the background." The way that works is that they
> have their own server that registers to your SIP port and when a call comes
> in their server generates a push to your iPhone and on receipt of that you
> open up the Softphone and somehow a handoff is done from their server to
> your phone. Just recently I contacted their tech support to find out what
> IP addresses I should use for their servers (they have multiple) so that I
> didn't have to open up 5061 to everyone. This is what I have in my custom
> rules...
>
>
>
> open_for_host_port()
>
> {
>
> local host="$1"
>
> local port="$2"
>
> local type="$3"
>
> local IFS=$'\n'
>
> local IPV4="$(host -t A $host)"
>
> local IPV6="$(host -t AAAA $host)"
>
> local IP=""
>
>
>
> for IP in $IPV4; do
>
> IP="${IP##* }"
>
> echo "[CUSTOM RULE] open port $port/$type for $host at $IP"
>
> ip4tables -A EXT_INPUT_CHAIN -s $IP -p $type -m $type --dport $port -j
> ACCEPT
>
> done
>
>
>
> for IP in $IPV6; do
>
> IP="${IP##* }"
>
> echo "[CUSTOM RULE] open port $port/$type for $host at $IP"
>
> ip6tables -A EXT_INPUT_CHAIN -s $IP -p $type -m $type --dport $port -j
> ACCEPT
>
> done
>
> }
>
> open_for_host_port "all.sipis.acrobits.cz" "5061" "tcp"
>
>
>
>
>
>
>
> On Sat, Nov 28, 2020 at 10:17 AM Lonnie Abelbeck <
> li...@lo...> wrote:
>
>
>
> > On Nov 28, 2020, at 5:17 AM, Michael Keuter <li...@mk...>
> wrote:
> >
> >>
> >> Am 27.11.2020 um 22:46 schrieb Michael Knill <
> mic...@ip...>:
> >>
> >> Hi Devs
> >>
> >> Hoping that I can get some advice from you in determining my
> development efforts moving forward.
> >> As driven by Covid, I have an increased number of customers that want
> mobile softphones and I have lost at least one customer and likely more
> because I have not been able to provide this functionality.
> >>
> >> I am looking to use Bria which I believe is the gold standard and I
> sort of have push notification working on it. The problem is that Push
> Notification work best with multiple registration which is not supported
> with chan_sip.
> >>
> >> So I have two options I believe:
> >> • Migrate to PJSIP
> >> • Add Kamailio to Astlinux to sip in front of Asterisk
> >>
> >> Now 1) will certainly be the easiest however 2) is something that I
> would love to do as its going to be more secure with external connecting
> clients.
> >>
> >> Any comments would be greatly appreciated?
> >>
> >> Regards
> >> Michael Knill
> >
> > Hi Michael,
> >
> > I have similar thoughts and would be interested in a more universal
> solution as well.
> >
> > I have one big customer, where I use WireGuard plus mostly Linphone
> without Push Notifications.
> > And the customer is quite happy with it, although they are not using it
> excessively.
> >
> > I created a (combined) template for an IP-phone plus a softphone under
> the same extension, same CLID, but with different SIP account names (with
> chan_sip).
> >
> > I generated the dialplan (in the template as well) so that both phones
> ring parallel.
> >
> > Most depends how the mobile OS handles priority for the SIP client on
> the phone and the the sleep functions.
> >
> > For Push Notifications I found this:
> >
> > https://www.zoiper.com/en/tutorials/push-notifications
> > https://github.com/balusreekanth/ios-asterisk-push
> >
> > BTW: For simple notifications (only for my own phone) I use Pushover
> (works also via email) or prepaid SMS via Voipbuster.
> >
> > https://pushover.net
> > https://www.voipbuster.com/sms_rates
> >
> > Michael
>
> Has anyone tested "Acrobits Groundwire" or "Acrobits Softphone" ?
>
> It looks like many "custom" mobile UC apps are based on the Acrobits SDK.
>
> Lonnie
>
>
>
>
>
> _______________________________________________
> Astlinux-devel mailing list
> Ast...@li...
> https://lists.sourceforge.net/lists/listinfo/astlinux-devel
>
> _______________________________________________
> Astlinux-devel mailing list
> Ast...@li...
> https://lists.sourceforge.net/lists/listinfo/astlinux-devel
>
> _______________________________________________
> Astlinux-devel mailing list
> Ast...@li...
> https://lists.sourceforge.net/lists/listinfo/astlinux-devel
>
|
|
From: David K. <da...@ke...> - 2020-12-19 22:18:22
|
Thanks Lonnie, that explains what I was seeing. I have implemented the background trick you suggest. Thanks David On Sat, Dec 19, 2020 at 10:13 AM Lonnie Abelbeck <li...@lo...> wrote: > > > On Dec 19, 2020, at 7:54 AM, David Kerr <da...@ke...> wrote: > > > > Can anyone tell me when DNS comes up during the boot process? I am > doing a host lookup within my firewall custom rules script $(host -t A > $fqdn) and it appears to get no IP address when the firewall is setup > during boot, but works fine if I do a firewall restart after the system has > booted up. > > > > Thanks > > Hi David, > > The order of the startup "Snn*" script symlinks are: > > pbx3 ~ # cd /etc/runlevels/default/ > > pbx3 ~ # ls -l S* > ... > lrwxrwxrwx 1 root root 20 Dec 16 21:34 S03network -> > ../../init.d/network > lrwxrwxrwx 1 root root 19 Dec 16 21:34 S04elocal -> > ../../init.d/elocal > lrwxrwxrwx 1 root root 21 Dec 16 21:34 S05watchdog -> > ../../init.d/watchdog > lrwxrwxrwx 1 root root 22 Dec 16 21:34 S06lmsensors -> > ../../init.d/lmsensors > lrwxrwxrwx 1 root root 19 Dec 16 21:34 S08msmtpd -> > ../../init.d/msmtpd > lrwxrwxrwx 1 root root 18 Dec 16 21:34 S09crond -> > ../../init.d/crond > lrwxrwxrwx 1 root root 21 Dec 16 21:34 S10iptables -> > ../../init.d/iptables > lrwxrwxrwx 1 root root 23 Dec 16 21:34 S11netsyslogd -> > ../../init.d/netsyslogd > lrwxrwxrwx 1 root root 18 Dec 16 21:34 S11snmpd -> > ../../init.d/snmpd > lrwxrwxrwx 1 root root 23 Dec 16 21:34 S12keepalived -> > ../../init.d/keepalived > lrwxrwxrwx 1 root root 21 Dec 16 21:34 S18dnscrypt -> > ../../init.d/dnscrypt > lrwxrwxrwx 1 root root 20 Dec 16 21:34 S18unbound -> > ../../init.d/unbound > lrwxrwxrwx 1 root root 20 Dec 16 21:34 S20dnsmasq -> > ../../init.d/dnsmasq > ... > > So the firewall starts at S10, and DNS starts at S20. > > If you want to keep the $(host -t A $fqdn) in the firewall custom-rules, > you could spawn a background sub-shell, something like: > > https://github.com/astlinux-project/astlinux/blob/c3ff8beba28533487c65d77e9ec2d54dd9facc3b/package/openssh/sshd.init#L144 > > Off the top of my head , untested... > -- > fqdn="example.com" > ( > cnt=6 > while [ $cnt -gt 0 ]; do > cnt=$((cnt - 1)) > sleep 10 > > A="$(host -t A $fqdn | sed -n -r -e 's#^.* has address > ([0-9.]+)$#\1#p')" > if [ -n "$A" ]; then > ## use $A example ## > echo "$A" > /tmp/test > exit > fi > done > ) >/dev/null 2>&1 & > -- > > Adjust the retry 'cnt' and sleep delay to your liking. > > Lonnie > > > > > _______________________________________________ > Astlinux-devel mailing list > Ast...@li... > https://lists.sourceforge.net/lists/listinfo/astlinux-devel > |
|
From: Lonnie A. <li...@lo...> - 2020-12-19 19:35:17
|
Announcing AstLinux Pre-Release: astlinux-1.4-4940-797713 ** Request for Testing This pre-release replaces getdns/stubby with unbound for DNS-over-TLS functionality. The latest Unbound 1.13.0 adds TCP/TLS stream reuse and TCP fast open which were perviously only supported in stubby. Additionally, unbound offers a cache. We have heard reports stubby would "hiccup" and hang at times, depending on the upstream provider. Please report any issues (or successes) with this change. ** IMPORTANT NOTICE -- The ipsec-tools (racoon) support in AstLinux is deprecated, and will be removed in the near future. The development of ipsec-tools has been ABANDONED. The Network tab -> VPN Type: "IPsec Peers" and "IPsec Mobile" is the web interface configuration for ipsec-tools. The AstLinux Team suggests using either WireGuard or OpenVPN for your VPN needs, but if IPsec is required for compatibility reasons the Network tab -> VPN Type: "IPsec strongSwan" should be able to do what you want, albeit in a less than ideal text based configuration. More info: https://doc.astlinux-project.org/userdoc:tt_ipsec_vpn_strongswan ** The AstLinux Team is regularly upgrading packages containing security and bug fixes as well as adding new features of our own. -- Linux Kernel 4.19.160 (version bump), security and bug fixes -- OpenSSL, version bump to 1.1.1i, security fix: CVE-2020-1971 -- WireGuard VPN, module 1.0.20201112 (version bump), tools 1.0.20200827 (no change) -- libcurl (curl) version bump to 7.74.0, security fixes: CVE-2020-8284, CVE-2020-8285, CVE-2020-8286 -- getdns/stubby, remove package, replace getdns/stubby with unbound for DNS-over-TLS -- unbound, version bump to 1.13.0 -- rng-tools, jitterentropy version bump to 3.0.0 (2x improved performance on APU systems) -- miniupnpd, version 2.1, add Debian security fixes: CVE-2019-12107, CVE-2019-12108 CVE-2019-12109, CVE-2019-12110, CVE-2019-12111 -- sngrep, version bump to 1.4.8 -- Monit, version bump to 5.27.1 -- zabbix, version bump to 4.0.27 -- Asterisk 13.29.2 ('13se' no change) Older than latest Asterisk 13.x version but more tested, built --without-pjproject -- Asterisk 13.38.0 (version bump) and 16.15.0 (version bump) -- DAHDI, dahdi-linux 3.1.0 (no change) and dahdi-tools 3.1.0 (no change) -- pjsip 2.10 (no change) -- Add support for directory /var/spool/asterisk/outgoing_tmp for call file staging. -- Add support for persistent /mnt/kd/call-file/ directory for certain tmpfs spool directories. If the directory /mnt/kd/call-file/ exists, the following symlinks will automatically occur: == /var/spool/asterisk/outgoing -> /mnt/kd/call-file/outgoing == /var/spool/asterisk/outgoing_tmp -> /mnt/kd/call-file/outgoing_tmp == /var/spool/asterisk/outgoing_done -> /mnt/kd/call-file/outgoing_done -- Complete Pre-Release ChangeLog: https://s3.amazonaws.com/beta.astlinux-project/astlinux-changelog/ChangeLog.txt The "AstLinux Pre-Release ChangeLog" and "Pre-Release Repository URL" entries can be found under the "Development" tab of the AstLinux Project web site ... AstLinux Project -> Development https://www.astlinux-project.org/dev.html AstLinux Team |
|
From: Lonnie A. <li...@lo...> - 2020-12-19 15:13:35
|
> On Dec 19, 2020, at 7:54 AM, David Kerr <da...@ke...> wrote: > > Can anyone tell me when DNS comes up during the boot process? I am doing a host lookup within my firewall custom rules script $(host -t A $fqdn) and it appears to get no IP address when the firewall is setup during boot, but works fine if I do a firewall restart after the system has booted up. > > Thanks Hi David, The order of the startup "Snn*" script symlinks are: pbx3 ~ # cd /etc/runlevels/default/ pbx3 ~ # ls -l S* ... lrwxrwxrwx 1 root root 20 Dec 16 21:34 S03network -> ../../init.d/network lrwxrwxrwx 1 root root 19 Dec 16 21:34 S04elocal -> ../../init.d/elocal lrwxrwxrwx 1 root root 21 Dec 16 21:34 S05watchdog -> ../../init.d/watchdog lrwxrwxrwx 1 root root 22 Dec 16 21:34 S06lmsensors -> ../../init.d/lmsensors lrwxrwxrwx 1 root root 19 Dec 16 21:34 S08msmtpd -> ../../init.d/msmtpd lrwxrwxrwx 1 root root 18 Dec 16 21:34 S09crond -> ../../init.d/crond lrwxrwxrwx 1 root root 21 Dec 16 21:34 S10iptables -> ../../init.d/iptables lrwxrwxrwx 1 root root 23 Dec 16 21:34 S11netsyslogd -> ../../init.d/netsyslogd lrwxrwxrwx 1 root root 18 Dec 16 21:34 S11snmpd -> ../../init.d/snmpd lrwxrwxrwx 1 root root 23 Dec 16 21:34 S12keepalived -> ../../init.d/keepalived lrwxrwxrwx 1 root root 21 Dec 16 21:34 S18dnscrypt -> ../../init.d/dnscrypt lrwxrwxrwx 1 root root 20 Dec 16 21:34 S18unbound -> ../../init.d/unbound lrwxrwxrwx 1 root root 20 Dec 16 21:34 S20dnsmasq -> ../../init.d/dnsmasq ... So the firewall starts at S10, and DNS starts at S20. If you want to keep the $(host -t A $fqdn) in the firewall custom-rules, you could spawn a background sub-shell, something like: https://github.com/astlinux-project/astlinux/blob/c3ff8beba28533487c65d77e9ec2d54dd9facc3b/package/openssh/sshd.init#L144 Off the top of my head , untested... -- fqdn="example.com" ( cnt=6 while [ $cnt -gt 0 ]; do cnt=$((cnt - 1)) sleep 10 A="$(host -t A $fqdn | sed -n -r -e 's#^.* has address ([0-9.]+)$#\1#p')" if [ -n "$A" ]; then ## use $A example ## echo "$A" > /tmp/test exit fi done ) >/dev/null 2>&1 & -- Adjust the retry 'cnt' and sleep delay to your liking. Lonnie |
|
From: David K. <da...@ke...> - 2020-12-19 14:00:00
|
Can anyone tell me when DNS comes up during the boot process? I am doing a host lookup within my firewall custom rules script $(host -t A $fqdn) and it appears to get no IP address when the firewall is setup during boot, but works fine if I do a firewall restart after the system has booted up. Thanks |
|
From: Lonnie A. <li...@lo...> - 2020-12-09 16:40:38
|
Thinking about this more, and Ramesh's testing, I bumped the RAM limit from 192 MB to 256 MB https://github.com/astlinux-project/astlinux/commit/1f8132fc3e51a146e4e8fc62849c8a996b5c9622 It is only a sanity limit, and does not effect the actual RAM usage. Lonnie > On Dec 7, 2020, at 6:59 PM, Ramesh GK <ram...@ho...> wrote: > > Thanks, Lonnie. > > Updating the MAX_IMAGEFS_SIZE resolved the problem and I have not seen any further issues. Thanks much for the quick response. > > Thanks > Ramesh GK > Phone: 408-462-0408 > > Strive not to be a success but rather to be of value - Albert Einstein > > From: Lonnie Abelbeck <li...@lo...> > Sent: Monday, December 7, 2020 12:15 PM > To: AstLinux Developers Mailing List <ast...@li...> > Subject: Re: [Astlinux-devel] Home Gateway Appliance - Issue with Free Space during Startup > > Hi David, > > We could ... but 192m gives us a little margin for the 256 MB FAT16 partition, and we have eaten into that margin recently. > > 1) Updating to Syslinux 6 (and memtest) used more space at the base. > > 2) As the size of RUNNIX increases, it uses more space, now about 9 MB each, we support a revert version, 18 MB total. > > 3) We now supply a unique initrd (astlinux-*.initrd.img) for every run image instead of just one shared, that added about 4 MB to the FAT16 partition. > > Currently our largest standard image is the Asterisk 16 build, which has a RAM based read-only filesystem of 137.2 MB, so the current limit of 192 MB offers significant headroom at the moment. > > Back to the 256 MB FAT16 partition needs: > > Syslinux 6 (and memtest): 0.5 MB > > RUNNIX (current): 9 MB > > RUNNIX (previous): 9 MB > > initrd (current): 4 MB > > initrd (previous): 4 MB > > So that leaves 229.5 MB to be shared by two compressed os/*.run images, a 256m value with 100 MB each max (compressed), yields less than 30 MB headroom, it would be getting tight, about only 10% free. The current 192m yields 74 MB each max (compressed), which gives us 81 MB of headroom out of the 256 MB FAT16 partition. > > I don't see us eating onto the 256 MB FAT16 partition other than the os/*.run images, but you never know. > > Good to do these calculations for time to time. > > Lonnie > > > > On Dec 7, 2020, at 9:35 AM, David Kerr <da...@ke...> wrote: > > > > Maybe we should just make the default build 256 MB. There can't be many systems now-a-days that could not accommodate that? > > > > David. > > > > On Mon, Dec 7, 2020 at 10:03 AM Lonnie Abelbeck <li...@lo...> wrote: > > > > > > > On Dec 7, 2020, at 8:18 AM, Lonnie Abelbeck <li...@lo...> wrote: > > > > > >> > > >> On Dec 6, 2020, at 10:54 PM, Ramesh GK <ram...@ho...> wrote: > > >> > > >> Hi, > > >> > > >> I started working on building some additions on top of astlinux 1.3.10 distribution without modifications to base packages and bumped into a strange issue after successful installation to the target system. > > >> > > >> During startup after loading the kernel, I am receiving the below message. > > >> > > >> Copying AstLinux files to RAM... > > >> ... > > >> ... > > >> ... > > >> cp: write error: no space left of device. > > >> cp: write error: no space left of device. > > >> cp: write error: no space left of device. > > >> cp: write error: no space left of device. > > >> cp: write error: no space left of device. > > >> > > >> After the above messages for some time it shows a bunch of errors and comes to a truncated login screen without any info that is normally shown at login prompt. > > >> > > >> I tried to increase the size during astlinux-makeimage creation from 256 to 512 which did not help. > > >> > > >> please let know if you have any ideas on how to resolve this issue. > > >> > > >> Thanks > > >> Ramesh GK > > > > > > By default, the upper limit to the RAM based read-only filesystem is 192 MB (uncompressed). > > > > > > This can be increased by bumping the MAX_IMAGEFS_SIZE variable in two places: > > > > > > https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fastlinux-project%2Fastlinux%2Fblob%2Fmaster%2Fscripts%2Fbuild%23L6&data=04%7C01%7C%7Cebb697e18fcd44d677fe08d89ad3cd35%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637429581755928094%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=4W4ajH4gMW5CG0sqFXVw1gD2bmx%2BOLxkRdVKrR08ixU%3D&reserved=0 > > > -- > > > MAX_IMAGEFS_SIZE=196608 # 192m > > > -- > > > > > > https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fastlinux-project%2Fastlinux%2Fblob%2Fmaster%2Fproject%2Finitrd%2Ftarget_skeleton%2Flinuxrc%23L6&data=04%7C01%7C%7Cebb697e18fcd44d677fe08d89ad3cd35%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637429581755928094%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=l6OoQH%2Bslcql1bvOeSThwjrtIOHucv85p%2FYKJ9KzhNE%3D&reserved=0 > > > -- > > > MAX_IMAGEFS_SIZE=196608 # 192m > > > -- > > > > > > Then, to rebuild the initrd, simply "rm initrd.img", and the next "./scripts/build/" will rebuild the initrd.img file. > > > > > > I would suggest 256m, which is 256*1024 = 262144 > > > -- > > > MAX_IMAGEFS_SIZE=262144 # 256m > > > -- > > > > > > Lonnie > > > > Replying to myself while I'm thinking about it ... > > > > The MAX_IMAGEFS_SIZE is also coupled with the size of the FAT16 partition size, 256 MB by default. > > > > We want to accommodate two (compressed) os/*.run images so we can revert-to-previous firmware. > > > > The uncompressed RAM size is around 2.6x the compressed size, so a 256m limit would result in approximately 100 MB compressed images, which should still work for two on a 256 MB FAT16 partition. > > > > Any higher value MAX_IMAGEFS_SIZE beyond 256m would require a larger FAT16 partition size if supporting two (compressed) os/*.run images is maintained. > > > > BTW, if no DAHDI hardware cards are used, setting BR2_PACKAGE_DAHDI_NO_CARD_FIRMWARE=y in .config will reduce the RAM based read-only filesystem by 8.8 MB. > > > > Lonnie > > > > > > > > > > > > > > > > > > > > _______________________________________________ > > Astlinux-devel mailing list > > Ast...@li... > > https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.sourceforge.net%2Flists%2Flistinfo%2Fastlinux-devel&data=04%7C01%7C%7Cebb697e18fcd44d677fe08d89ad3cd35%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637429581755928094%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=z7LPJgzXTB5%2FarC67WsQdAzMhVAAj%2BpTSRXHZ%2FN7v7c%3D&reserved=0 > > _______________________________________________ > > Astlinux-devel mailing list > > Ast...@li... > > https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.sourceforge.net%2Flists%2Flistinfo%2Fastlinux-devel&data=04%7C01%7C%7Cebb697e18fcd44d677fe08d89ad3cd35%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637429581755928094%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=z7LPJgzXTB5%2FarC67WsQdAzMhVAAj%2BpTSRXHZ%2FN7v7c%3D&reserved=0 > > > > _______________________________________________ > Astlinux-devel mailing list > Ast...@li... > https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.sourceforge.net%2Flists%2Flistinfo%2Fastlinux-devel&data=04%7C01%7C%7Cebb697e18fcd44d677fe08d89ad3cd35%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637429581755938089%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=V21vO%2Be90tODE0Sd%2BIOIj8j2J08imOYOqY5h8dm1n%2B0%3D&reserved=0 > _______________________________________________ > Astlinux-devel mailing list > Ast...@li... > https://lists.sourceforge.net/lists/listinfo/astlinux-devel |
|
From: Ramesh GK <ram...@ho...> - 2020-12-08 00:59:49
|
Thanks, Lonnie. Updating the MAX_IMAGEFS_SIZE resolved the problem and I have not seen any further issues. Thanks much for the quick response. Thanks Ramesh GK Phone: 408-462-0408 Strive not to be a success but rather to be of value - Albert Einstein ________________________________ From: Lonnie Abelbeck <li...@lo...> Sent: Monday, December 7, 2020 12:15 PM To: AstLinux Developers Mailing List <ast...@li...> Subject: Re: [Astlinux-devel] Home Gateway Appliance - Issue with Free Space during Startup Hi David, We could ... but 192m gives us a little margin for the 256 MB FAT16 partition, and we have eaten into that margin recently. 1) Updating to Syslinux 6 (and memtest) used more space at the base. 2) As the size of RUNNIX increases, it uses more space, now about 9 MB each, we support a revert version, 18 MB total. 3) We now supply a unique initrd (astlinux-*.initrd.img) for every run image instead of just one shared, that added about 4 MB to the FAT16 partition. Currently our largest standard image is the Asterisk 16 build, which has a RAM based read-only filesystem of 137.2 MB, so the current limit of 192 MB offers significant headroom at the moment. Back to the 256 MB FAT16 partition needs: Syslinux 6 (and memtest): 0.5 MB RUNNIX (current): 9 MB RUNNIX (previous): 9 MB initrd (current): 4 MB initrd (previous): 4 MB So that leaves 229.5 MB to be shared by two compressed os/*.run images, a 256m value with 100 MB each max (compressed), yields less than 30 MB headroom, it would be getting tight, about only 10% free. The current 192m yields 74 MB each max (compressed), which gives us 81 MB of headroom out of the 256 MB FAT16 partition. I don't see us eating onto the 256 MB FAT16 partition other than the os/*.run images, but you never know. Good to do these calculations for time to time. Lonnie > On Dec 7, 2020, at 9:35 AM, David Kerr <da...@ke...> wrote: > > Maybe we should just make the default build 256 MB. There can't be many systems now-a-days that could not accommodate that? > > David. > > On Mon, Dec 7, 2020 at 10:03 AM Lonnie Abelbeck <li...@lo...> wrote: > > > > On Dec 7, 2020, at 8:18 AM, Lonnie Abelbeck <li...@lo...> wrote: > > > >> > >> On Dec 6, 2020, at 10:54 PM, Ramesh GK <ram...@ho...> wrote: > >> > >> Hi, > >> > >> I started working on building some additions on top of astlinux 1.3.10 distribution without modifications to base packages and bumped into a strange issue after successful installation to the target system. > >> > >> During startup after loading the kernel, I am receiving the below message. > >> > >> Copying AstLinux files to RAM... > >> ... > >> ... > >> ... > >> cp: write error: no space left of device. > >> cp: write error: no space left of device. > >> cp: write error: no space left of device. > >> cp: write error: no space left of device. > >> cp: write error: no space left of device. > >> > >> After the above messages for some time it shows a bunch of errors and comes to a truncated login screen without any info that is normally shown at login prompt. > >> > >> I tried to increase the size during astlinux-makeimage creation from 256 to 512 which did not help. > >> > >> please let know if you have any ideas on how to resolve this issue. > >> > >> Thanks > >> Ramesh GK > > > > By default, the upper limit to the RAM based read-only filesystem is 192 MB (uncompressed). > > > > This can be increased by bumping the MAX_IMAGEFS_SIZE variable in two places: > > > > https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fastlinux-project%2Fastlinux%2Fblob%2Fmaster%2Fscripts%2Fbuild%23L6&data=04%7C01%7C%7Cebb697e18fcd44d677fe08d89ad3cd35%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637429581755928094%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=4W4ajH4gMW5CG0sqFXVw1gD2bmx%2BOLxkRdVKrR08ixU%3D&reserved=0 > > -- > > MAX_IMAGEFS_SIZE=196608 # 192m > > -- > > > > https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fastlinux-project%2Fastlinux%2Fblob%2Fmaster%2Fproject%2Finitrd%2Ftarget_skeleton%2Flinuxrc%23L6&data=04%7C01%7C%7Cebb697e18fcd44d677fe08d89ad3cd35%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637429581755928094%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=l6OoQH%2Bslcql1bvOeSThwjrtIOHucv85p%2FYKJ9KzhNE%3D&reserved=0 > > -- > > MAX_IMAGEFS_SIZE=196608 # 192m > > -- > > > > Then, to rebuild the initrd, simply "rm initrd.img", and the next "./scripts/build/" will rebuild the initrd.img file. > > > > I would suggest 256m, which is 256*1024 = 262144 > > -- > > MAX_IMAGEFS_SIZE=262144 # 256m > > -- > > > > Lonnie > > Replying to myself while I'm thinking about it ... > > The MAX_IMAGEFS_SIZE is also coupled with the size of the FAT16 partition size, 256 MB by default. > > We want to accommodate two (compressed) os/*.run images so we can revert-to-previous firmware. > > The uncompressed RAM size is around 2.6x the compressed size, so a 256m limit would result in approximately 100 MB compressed images, which should still work for two on a 256 MB FAT16 partition. > > Any higher value MAX_IMAGEFS_SIZE beyond 256m would require a larger FAT16 partition size if supporting two (compressed) os/*.run images is maintained. > > BTW, if no DAHDI hardware cards are used, setting BR2_PACKAGE_DAHDI_NO_CARD_FIRMWARE=y in .config will reduce the RAM based read-only filesystem by 8.8 MB. > > Lonnie > > > > > > > > > > _______________________________________________ > Astlinux-devel mailing list > Ast...@li... > https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.sourceforge.net%2Flists%2Flistinfo%2Fastlinux-devel&data=04%7C01%7C%7Cebb697e18fcd44d677fe08d89ad3cd35%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637429581755928094%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=z7LPJgzXTB5%2FarC67WsQdAzMhVAAj%2BpTSRXHZ%2FN7v7c%3D&reserved=0 > _______________________________________________ > Astlinux-devel mailing list > Ast...@li... > https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.sourceforge.net%2Flists%2Flistinfo%2Fastlinux-devel&data=04%7C01%7C%7Cebb697e18fcd44d677fe08d89ad3cd35%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637429581755928094%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=z7LPJgzXTB5%2FarC67WsQdAzMhVAAj%2BpTSRXHZ%2FN7v7c%3D&reserved=0 _______________________________________________ Astlinux-devel mailing list Ast...@li... https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.sourceforge.net%2Flists%2Flistinfo%2Fastlinux-devel&data=04%7C01%7C%7Cebb697e18fcd44d677fe08d89ad3cd35%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637429581755938089%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=V21vO%2Be90tODE0Sd%2BIOIj8j2J08imOYOqY5h8dm1n%2B0%3D&reserved=0 |
|
From: Lonnie A. <li...@lo...> - 2020-12-07 17:16:00
|
Hi David, We could ... but 192m gives us a little margin for the 256 MB FAT16 partition, and we have eaten into that margin recently. 1) Updating to Syslinux 6 (and memtest) used more space at the base. 2) As the size of RUNNIX increases, it uses more space, now about 9 MB each, we support a revert version, 18 MB total. 3) We now supply a unique initrd (astlinux-*.initrd.img) for every run image instead of just one shared, that added about 4 MB to the FAT16 partition. Currently our largest standard image is the Asterisk 16 build, which has a RAM based read-only filesystem of 137.2 MB, so the current limit of 192 MB offers significant headroom at the moment. Back to the 256 MB FAT16 partition needs: Syslinux 6 (and memtest): 0.5 MB RUNNIX (current): 9 MB RUNNIX (previous): 9 MB initrd (current): 4 MB initrd (previous): 4 MB So that leaves 229.5 MB to be shared by two compressed os/*.run images, a 256m value with 100 MB each max (compressed), yields less than 30 MB headroom, it would be getting tight, about only 10% free. The current 192m yields 74 MB each max (compressed), which gives us 81 MB of headroom out of the 256 MB FAT16 partition. I don't see us eating onto the 256 MB FAT16 partition other than the os/*.run images, but you never know. Good to do these calculations for time to time. Lonnie > On Dec 7, 2020, at 9:35 AM, David Kerr <da...@ke...> wrote: > > Maybe we should just make the default build 256 MB. There can't be many systems now-a-days that could not accommodate that? > > David. > > On Mon, Dec 7, 2020 at 10:03 AM Lonnie Abelbeck <li...@lo...> wrote: > > > > On Dec 7, 2020, at 8:18 AM, Lonnie Abelbeck <li...@lo...> wrote: > > > >> > >> On Dec 6, 2020, at 10:54 PM, Ramesh GK <ram...@ho...> wrote: > >> > >> Hi, > >> > >> I started working on building some additions on top of astlinux 1.3.10 distribution without modifications to base packages and bumped into a strange issue after successful installation to the target system. > >> > >> During startup after loading the kernel, I am receiving the below message. > >> > >> Copying AstLinux files to RAM... > >> ... > >> ... > >> ... > >> cp: write error: no space left of device. > >> cp: write error: no space left of device. > >> cp: write error: no space left of device. > >> cp: write error: no space left of device. > >> cp: write error: no space left of device. > >> > >> After the above messages for some time it shows a bunch of errors and comes to a truncated login screen without any info that is normally shown at login prompt. > >> > >> I tried to increase the size during astlinux-makeimage creation from 256 to 512 which did not help. > >> > >> please let know if you have any ideas on how to resolve this issue. > >> > >> Thanks > >> Ramesh GK > > > > By default, the upper limit to the RAM based read-only filesystem is 192 MB (uncompressed). > > > > This can be increased by bumping the MAX_IMAGEFS_SIZE variable in two places: > > > > https://github.com/astlinux-project/astlinux/blob/master/scripts/build#L6 > > -- > > MAX_IMAGEFS_SIZE=196608 # 192m > > -- > > > > https://github.com/astlinux-project/astlinux/blob/master/project/initrd/target_skeleton/linuxrc#L6 > > -- > > MAX_IMAGEFS_SIZE=196608 # 192m > > -- > > > > Then, to rebuild the initrd, simply "rm initrd.img", and the next "./scripts/build/" will rebuild the initrd.img file. > > > > I would suggest 256m, which is 256*1024 = 262144 > > -- > > MAX_IMAGEFS_SIZE=262144 # 256m > > -- > > > > Lonnie > > Replying to myself while I'm thinking about it ... > > The MAX_IMAGEFS_SIZE is also coupled with the size of the FAT16 partition size, 256 MB by default. > > We want to accommodate two (compressed) os/*.run images so we can revert-to-previous firmware. > > The uncompressed RAM size is around 2.6x the compressed size, so a 256m limit would result in approximately 100 MB compressed images, which should still work for two on a 256 MB FAT16 partition. > > Any higher value MAX_IMAGEFS_SIZE beyond 256m would require a larger FAT16 partition size if supporting two (compressed) os/*.run images is maintained. > > BTW, if no DAHDI hardware cards are used, setting BR2_PACKAGE_DAHDI_NO_CARD_FIRMWARE=y in .config will reduce the RAM based read-only filesystem by 8.8 MB. > > Lonnie > > > > > > > > > > _______________________________________________ > Astlinux-devel mailing list > Ast...@li... > https://lists.sourceforge.net/lists/listinfo/astlinux-devel > _______________________________________________ > Astlinux-devel mailing list > Ast...@li... > https://lists.sourceforge.net/lists/listinfo/astlinux-devel |
|
From: David K. <da...@ke...> - 2020-12-07 16:06:47
|
Maybe we should just make the default build 256 MB. There can't be many systems now-a-days that could not accommodate that? David. On Mon, Dec 7, 2020 at 10:03 AM Lonnie Abelbeck <li...@lo...> wrote: > > > > On Dec 7, 2020, at 8:18 AM, Lonnie Abelbeck <li...@lo...> > wrote: > > > >> > >> On Dec 6, 2020, at 10:54 PM, Ramesh GK <ram...@ho...> wrote: > >> > >> Hi, > >> > >> I started working on building some additions on top of astlinux 1.3.10 > distribution without modifications to base packages and bumped into a > strange issue after successful installation to the target system. > >> > >> During startup after loading the kernel, I am receiving the below > message. > >> > >> Copying AstLinux files to RAM... > >> ... > >> ... > >> ... > >> cp: write error: no space left of device. > >> cp: write error: no space left of device. > >> cp: write error: no space left of device. > >> cp: write error: no space left of device. > >> cp: write error: no space left of device. > >> > >> After the above messages for some time it shows a bunch of errors and > comes to a truncated login screen without any info that is normally shown > at login prompt. > >> > >> I tried to increase the size during astlinux-makeimage creation from > 256 to 512 which did not help. > >> > >> please let know if you have any ideas on how to resolve this issue. > >> > >> Thanks > >> Ramesh GK > > > > By default, the upper limit to the RAM based read-only filesystem is 192 > MB (uncompressed). > > > > This can be increased by bumping the MAX_IMAGEFS_SIZE variable in two > places: > > > > > https://github.com/astlinux-project/astlinux/blob/master/scripts/build#L6 > > -- > > MAX_IMAGEFS_SIZE=196608 # 192m > > -- > > > > > https://github.com/astlinux-project/astlinux/blob/master/project/initrd/target_skeleton/linuxrc#L6 > > -- > > MAX_IMAGEFS_SIZE=196608 # 192m > > -- > > > > Then, to rebuild the initrd, simply "rm initrd.img", and the next > "./scripts/build/" will rebuild the initrd.img file. > > > > I would suggest 256m, which is 256*1024 = 262144 > > -- > > MAX_IMAGEFS_SIZE=262144 # 256m > > -- > > > > Lonnie > > Replying to myself while I'm thinking about it ... > > The MAX_IMAGEFS_SIZE is also coupled with the size of the FAT16 partition > size, 256 MB by default. > > We want to accommodate two (compressed) os/*.run images so we can > revert-to-previous firmware. > > The uncompressed RAM size is around 2.6x the compressed size, so a 256m > limit would result in approximately 100 MB compressed images, which should > still work for two on a 256 MB FAT16 partition. > > Any higher value MAX_IMAGEFS_SIZE beyond 256m would require a larger FAT16 > partition size if supporting two (compressed) os/*.run images is maintained. > > BTW, if no DAHDI hardware cards are used, setting > BR2_PACKAGE_DAHDI_NO_CARD_FIRMWARE=y in .config will reduce the RAM based > read-only filesystem by 8.8 MB. > > Lonnie > > > > > > > > > > _______________________________________________ > Astlinux-devel mailing list > Ast...@li... > https://lists.sourceforge.net/lists/listinfo/astlinux-devel > |
|
From: Lonnie A. <li...@lo...> - 2020-12-07 15:03:41
|
> On Dec 7, 2020, at 8:18 AM, Lonnie Abelbeck <li...@lo...> wrote: > >> >> On Dec 6, 2020, at 10:54 PM, Ramesh GK <ram...@ho...> wrote: >> >> Hi, >> >> I started working on building some additions on top of astlinux 1.3.10 distribution without modifications to base packages and bumped into a strange issue after successful installation to the target system. >> >> During startup after loading the kernel, I am receiving the below message. >> >> Copying AstLinux files to RAM... >> ... >> ... >> ... >> cp: write error: no space left of device. >> cp: write error: no space left of device. >> cp: write error: no space left of device. >> cp: write error: no space left of device. >> cp: write error: no space left of device. >> >> After the above messages for some time it shows a bunch of errors and comes to a truncated login screen without any info that is normally shown at login prompt. >> >> I tried to increase the size during astlinux-makeimage creation from 256 to 512 which did not help. >> >> please let know if you have any ideas on how to resolve this issue. >> >> Thanks >> Ramesh GK > > By default, the upper limit to the RAM based read-only filesystem is 192 MB (uncompressed). > > This can be increased by bumping the MAX_IMAGEFS_SIZE variable in two places: > > https://github.com/astlinux-project/astlinux/blob/master/scripts/build#L6 > -- > MAX_IMAGEFS_SIZE=196608 # 192m > -- > > https://github.com/astlinux-project/astlinux/blob/master/project/initrd/target_skeleton/linuxrc#L6 > -- > MAX_IMAGEFS_SIZE=196608 # 192m > -- > > Then, to rebuild the initrd, simply "rm initrd.img", and the next "./scripts/build/" will rebuild the initrd.img file. > > I would suggest 256m, which is 256*1024 = 262144 > -- > MAX_IMAGEFS_SIZE=262144 # 256m > -- > > Lonnie Replying to myself while I'm thinking about it ... The MAX_IMAGEFS_SIZE is also coupled with the size of the FAT16 partition size, 256 MB by default. We want to accommodate two (compressed) os/*.run images so we can revert-to-previous firmware. The uncompressed RAM size is around 2.6x the compressed size, so a 256m limit would result in approximately 100 MB compressed images, which should still work for two on a 256 MB FAT16 partition. Any higher value MAX_IMAGEFS_SIZE beyond 256m would require a larger FAT16 partition size if supporting two (compressed) os/*.run images is maintained. BTW, if no DAHDI hardware cards are used, setting BR2_PACKAGE_DAHDI_NO_CARD_FIRMWARE=y in .config will reduce the RAM based read-only filesystem by 8.8 MB. Lonnie |
|
From: Lonnie A. <li...@lo...> - 2020-12-07 14:18:41
|
> On Dec 6, 2020, at 10:54 PM, Ramesh GK <ram...@ho...> wrote: > > Hi, > > I started working on building some additions on top of astlinux 1.3.10 distribution without modifications to base packages and bumped into a strange issue after successful installation to the target system. > > During startup after loading the kernel, I am receiving the below message. > > Copying AstLinux files to RAM... > ... > ... > ... > cp: write error: no space left of device. > cp: write error: no space left of device. > cp: write error: no space left of device. > cp: write error: no space left of device. > cp: write error: no space left of device. > > After the above messages for some time it shows a bunch of errors and comes to a truncated login screen without any info that is normally shown at login prompt. > > I tried to increase the size during astlinux-makeimage creation from 256 to 512 which did not help. > > please let know if you have any ideas on how to resolve this issue. > > Thanks > Ramesh GK By default, the upper limit to the RAM based read-only filesystem is 192 MB (uncompressed). This can be increased by bumping the MAX_IMAGEFS_SIZE variable in two places: https://github.com/astlinux-project/astlinux/blob/master/scripts/build#L6 -- MAX_IMAGEFS_SIZE=196608 # 192m -- https://github.com/astlinux-project/astlinux/blob/master/project/initrd/target_skeleton/linuxrc#L6 -- MAX_IMAGEFS_SIZE=196608 # 192m -- Then, to rebuild the initrd, simply "rm initrd.img", and the next "./scripts/build/" will rebuild the initrd.img file. I would suggest 256m, which is 256*1024 = 262144 -- MAX_IMAGEFS_SIZE=262144 # 256m -- Lonnie |
|
From: Ramesh GK <ram...@ho...> - 2020-12-07 04:55:07
|
Hi, I started working on building some additions on top of astlinux 1.3.10 distribution without modifications to base packages and bumped into a strange issue after successful installation to the target system. During startup after loading the kernel, I am receiving the below message. Copying AstLinux files to RAM... ... ... ... cp: write error: no space left of device. cp: write error: no space left of device. cp: write error: no space left of device. cp: write error: no space left of device. cp: write error: no space left of device. After the above messages for some time it shows a bunch of errors and comes to a truncated login screen without any info that is normally shown at login prompt. I tried to increase the size during astlinux-makeimage creation from 256 to 512 which did not help. please let know if you have any ideas on how to resolve this issue. Thanks Ramesh GK Strive not to be a success but rather to be of value - Albert Einstein |
|
From: Michael K. <mic...@ip...> - 2020-12-04 21:28:44
|
I do have it set up at one site but more than happy to not support it as its giving me grief.
Wireguard and OpenVPN only moving forward. As Mikrotik will be supporting Wireguard soon it's a no brainer.
Regards
Michael Knill
On 5/12/20, 7:20 am, "Lonnie Abelbeck" <li...@lo...> wrote:
Hi AstLinux Users and Devs,
Ever since AstLinux version 0.6.0 (2008-08-30) the IPsec VPN was supported using the ipsec-tools (racoon) package.
To this day, the Network tab -> VPN Type: "IPsec Peers" and "IPsec Mobile" supports a web interface configuration for ipsec-tools.
Notice: The ipsec-tools (racoon) support in AstLinux is deprecated, and will be removed in the near future.
As an alternative solution, AstLinux IPsec VPN users can use the Network tab -> VPN Type: "IPsec strongSwan"
This email is to give notice that VPN Type: "IPsec Peers" and "IPsec Mobile" support is deprecated, also the AstLinux Team would like to know how common this implementation is being used today.
Please respond to this email if you are using the VPN Type: "IPsec Peers" and/or "IPsec Mobile".
Justification:
A) "The development of ipsec-tools has been ABANDONED." [1]
B) Buidroot has removed the ipsec-tools package [2]
C) Debian 10 (buster) no longer includes ipsec-tools (racoon).
The AstLinux Team suggests using either WireGuard or OpenVPN for your VPN needs, but if IPsec is required for compatibility reasons the Network tab -> VPN Type: "IPsec strongSwan" should be able to do what you want, albeit in a less than ideal text based configuration.
More Info: IPsec VPN (strongSwan) Configuration
https://doc.astlinux.org/userdoc:tt_ipsec_vpn_strongswan
AstLinux Team
[1] http://ipsec-tools.sourceforge.net/
[2] https://git.busybox.net/buildroot/commit/?id=a3f58a74e0094bbf960ca9f4161c492b00a849ef
_______________________________________________
Astlinux-devel mailing list
Ast...@li...
https://lists.sourceforge.net/lists/listinfo/astlinux-devel
|
|
From: Lonnie A. <li...@lo...> - 2020-12-04 20:20:28
|
Hi AstLinux Users and Devs, Ever since AstLinux version 0.6.0 (2008-08-30) the IPsec VPN was supported using the ipsec-tools (racoon) package. To this day, the Network tab -> VPN Type: "IPsec Peers" and "IPsec Mobile" supports a web interface configuration for ipsec-tools. Notice: The ipsec-tools (racoon) support in AstLinux is deprecated, and will be removed in the near future. As an alternative solution, AstLinux IPsec VPN users can use the Network tab -> VPN Type: "IPsec strongSwan" This email is to give notice that VPN Type: "IPsec Peers" and "IPsec Mobile" support is deprecated, also the AstLinux Team would like to know how common this implementation is being used today. Please respond to this email if you are using the VPN Type: "IPsec Peers" and/or "IPsec Mobile". Justification: A) "The development of ipsec-tools has been ABANDONED." [1] B) Buidroot has removed the ipsec-tools package [2] C) Debian 10 (buster) no longer includes ipsec-tools (racoon). The AstLinux Team suggests using either WireGuard or OpenVPN for your VPN needs, but if IPsec is required for compatibility reasons the Network tab -> VPN Type: "IPsec strongSwan" should be able to do what you want, albeit in a less than ideal text based configuration. More Info: IPsec VPN (strongSwan) Configuration https://doc.astlinux.org/userdoc:tt_ipsec_vpn_strongswan AstLinux Team [1] http://ipsec-tools.sourceforge.net/ [2] https://git.busybox.net/buildroot/commit/?id=a3f58a74e0094bbf960ca9f4161c492b00a849ef |
|
From: Michael K. <mic...@ip...> - 2020-11-28 23:42:13
|
Thanks David Yes I have used Bria with Wireguard VPN and it works fine. Also fails over well when roaming between Wifi and 4G. Yes you need to manually activate it which is annoying and I think it may prevent the softphone from backgrounding so battery problems as well. Interestingly however, I did some testing this morning with my Acrobits portal and the Cloud Softphone product on a test account using TLS. Provisioning was a matter of pointing the phone to a QR code and it worked first time. There appeared to be no issues with re-registration between the push server and direct and I did a test driving down the road and there was a short break but it roamed successfully between Wifi and 4G. Bit annoying as the call only stays up for 1m so not really a long enough test but long enough to notice successful roaming. I think its time to test it for real now. I will keep you posted. Thanks all for the feedback. Regards Michael Knill From: David Kerr <da...@ke...> Reply to: AstLinux Developers Mailing List <ast...@li...> Date: Sunday, 29 November 2020 at 9:51 am To: AstLinux Developers Mailing List <ast...@li...> Subject: Re: [Astlinux-devel] Recommendations for my architecture moving forward Looking at https://developer.apple.com/documentation/networkextension/personal_vpn/vpn_on_demand_rules it seems that Wireguard have implemented very basic VPN on-demand (turn it on when on cellular or when on wifi). What would be needed is much more granular control, at a minimum turn on for specific applications only. Even better would be connection specific rules. That feels like a lot of work which would have to get implemented in the Wireguard iOS app. David On Sat, Nov 28, 2020 at 5:35 PM David Kerr <da...@ke...<mailto:da...@ke...>> wrote: I used to have 5061 open to anything, but I started to get attempts to login and make calls, so I now restrict it to the Acrobits servers. For the iPhone itself I have to be connected to my home network either directly to the WiFi or by VPN (I use Wireguard). Acrobits (afaik) has no built-in VPN. I have not tried the Wireguard on-demand setting but it is possible that might work. I have not tried roaming between WiFi and Mobile. David On Sat, Nov 28, 2020 at 3:44 PM Michael Knill <mic...@ip...<mailto:mic...@ip...>> wrote: Hi All Thanks for responding. Bria is certainly the gold standard however I have been down the Acrobits path and had a full demo of what they can provide. They have a full white labelling solution and there provisioning system is pretty cool. Just point your phone to a QR code. David are you only opening up 5061 to the push server? If so then your iPhone must never directly register to your system externally? Also another question; can you roam between Wi-Fi and 4G while on a call? Regards Michael Knill From: David Kerr <da...@ke...<mailto:da...@ke...>> Reply to: AstLinux Developers Mailing List <ast...@li...<mailto:ast...@li...>> Date: Sunday, 29 November 2020 at 2:48 am To: AstLinux Developers Mailing List <ast...@li...<mailto:ast...@li...>> Subject: Re: [Astlinux-devel] Recommendations for my architecture moving forward I have had Acrobits Softphone installed on my iPhone for several years. I don't use it a whole lot but it works well. It has push notification for when the softphone is "in the background." The way that works is that they have their own server that registers to your SIP port and when a call comes in their server generates a push to your iPhone and on receipt of that you open up the Softphone and somehow a handoff is done from their server to your phone. Just recently I contacted their tech support to find out what IP addresses I should use for their servers (they have multiple) so that I didn't have to open up 5061 to everyone. This is what I have in my custom rules... open_for_host_port() { local host="$1" local port="$2" local type="$3" local IFS=$'\n' local IPV4="$(host -t A $host)" local IPV6="$(host -t AAAA $host)" local IP="" for IP in $IPV4; do IP="${IP##* }" echo "[CUSTOM RULE] open port $port/$type for $host at $IP" ip4tables -A EXT_INPUT_CHAIN -s $IP -p $type -m $type --dport $port -j ACCEPT done for IP in $IPV6; do IP="${IP##* }" echo "[CUSTOM RULE] open port $port/$type for $host at $IP" ip6tables -A EXT_INPUT_CHAIN -s $IP -p $type -m $type --dport $port -j ACCEPT done } open_for_host_port "all.sipis.acrobits.cz<http://all.sipis.acrobits.cz>" "5061" "tcp" On Sat, Nov 28, 2020 at 10:17 AM Lonnie Abelbeck <li...@lo...<mailto:li...@lo...>> wrote: > On Nov 28, 2020, at 5:17 AM, Michael Keuter <li...@mk...<mailto:li...@mk...>> wrote: > >> >> Am 27.11.2020 um 22:46 schrieb Michael Knill <mic...@ip...<mailto:mic...@ip...>>: >> >> Hi Devs >> >> Hoping that I can get some advice from you in determining my development efforts moving forward. >> As driven by Covid, I have an increased number of customers that want mobile softphones and I have lost at least one customer and likely more because I have not been able to provide this functionality. >> >> I am looking to use Bria which I believe is the gold standard and I sort of have push notification working on it. The problem is that Push Notification work best with multiple registration which is not supported with chan_sip. >> >> So I have two options I believe: >> • Migrate to PJSIP >> • Add Kamailio to Astlinux to sip in front of Asterisk >> >> Now 1) will certainly be the easiest however 2) is something that I would love to do as its going to be more secure with external connecting clients. >> >> Any comments would be greatly appreciated? >> >> Regards >> Michael Knill > > Hi Michael, > > I have similar thoughts and would be interested in a more universal solution as well. > > I have one big customer, where I use WireGuard plus mostly Linphone without Push Notifications. > And the customer is quite happy with it, although they are not using it excessively. > > I created a (combined) template for an IP-phone plus a softphone under the same extension, same CLID, but with different SIP account names (with chan_sip). > > I generated the dialplan (in the template as well) so that both phones ring parallel. > > Most depends how the mobile OS handles priority for the SIP client on the phone and the the sleep functions. > > For Push Notifications I found this: > > https://www.zoiper.com/en/tutorials/push-notifications > https://github.com/balusreekanth/ios-asterisk-push > > BTW: For simple notifications (only for my own phone) I use Pushover (works also via email) or prepaid SMS via Voipbuster. > > https://pushover.net > https://www.voipbuster.com/sms_rates > > Michael Has anyone tested "Acrobits Groundwire" or "Acrobits Softphone" ? It looks like many "custom" mobile UC apps are based on the Acrobits SDK. Lonnie _______________________________________________ Astlinux-devel mailing list Ast...@li...<mailto:Ast...@li...> https://lists.sourceforge.net/lists/listinfo/astlinux-devel _______________________________________________ Astlinux-devel mailing list Ast...@li...<mailto:Ast...@li...> https://lists.sourceforge.net/lists/listinfo/astlinux-devel |
|
From: David K. <da...@ke...> - 2020-11-28 23:24:30
|
I used to have 5061 open to anything, but I started to get attempts to
login and make calls, so I now restrict it to the Acrobits servers. For
the iPhone itself I have to be connected to my home network either directly
to the WiFi or by VPN (I use Wireguard). Acrobits (afaik) has no built-in
VPN. I have not tried the Wireguard on-demand setting but it is possible
that might work.
I have not tried roaming between WiFi and Mobile.
David
On Sat, Nov 28, 2020 at 3:44 PM Michael Knill <
mic...@ip...> wrote:
> Hi All
>
>
>
> Thanks for responding.
>
> Bria is certainly the gold standard however I have been down the Acrobits
> path and had a full demo of what they can provide. They have a full white
> labelling solution and there provisioning system is pretty cool. Just point
> your phone to a QR code.
>
>
>
> David are you only opening up 5061 to the push server? If so then your
> iPhone must never directly register to your system externally?
>
> Also another question; can you roam between Wi-Fi and 4G while on a call?
>
>
>
> Regards
>
> Michael Knill
>
>
>
> *From: *David Kerr <da...@ke...>
> *Reply to: *AstLinux Developers Mailing List <
> ast...@li...>
> *Date: *Sunday, 29 November 2020 at 2:48 am
> *To: *AstLinux Developers Mailing List <
> ast...@li...>
> *Subject: *Re: [Astlinux-devel] Recommendations for my architecture
> moving forward
>
>
>
> I have had Acrobits Softphone installed on my iPhone for several years. I
> don't use it a whole lot but it works well. It has push notification for
> when the softphone is "in the background." The way that works is that they
> have their own server that registers to your SIP port and when a call comes
> in their server generates a push to your iPhone and on receipt of that you
> open up the Softphone and somehow a handoff is done from their server to
> your phone. Just recently I contacted their tech support to find out what
> IP addresses I should use for their servers (they have multiple) so that I
> didn't have to open up 5061 to everyone. This is what I have in my custom
> rules...
>
>
>
> open_for_host_port()
>
> {
>
> local host="$1"
>
> local port="$2"
>
> local type="$3"
>
> local IFS=$'\n'
>
> local IPV4="$(host -t A $host)"
>
> local IPV6="$(host -t AAAA $host)"
>
> local IP=""
>
>
>
> for IP in $IPV4; do
>
> IP="${IP##* }"
>
> echo "[CUSTOM RULE] open port $port/$type for $host at $IP"
>
> ip4tables -A EXT_INPUT_CHAIN -s $IP -p $type -m $type --dport $port -j
> ACCEPT
>
> done
>
>
>
> for IP in $IPV6; do
>
> IP="${IP##* }"
>
> echo "[CUSTOM RULE] open port $port/$type for $host at $IP"
>
> ip6tables -A EXT_INPUT_CHAIN -s $IP -p $type -m $type --dport $port -j
> ACCEPT
>
> done
>
> }
>
> open_for_host_port "all.sipis.acrobits.cz" "5061" "tcp"
>
>
>
>
>
>
>
> On Sat, Nov 28, 2020 at 10:17 AM Lonnie Abelbeck <
> li...@lo...> wrote:
>
>
>
> > On Nov 28, 2020, at 5:17 AM, Michael Keuter <li...@mk...>
> wrote:
> >
> >>
> >> Am 27.11.2020 um 22:46 schrieb Michael Knill <
> mic...@ip...>:
> >>
> >> Hi Devs
> >>
> >> Hoping that I can get some advice from you in determining my
> development efforts moving forward.
> >> As driven by Covid, I have an increased number of customers that want
> mobile softphones and I have lost at least one customer and likely more
> because I have not been able to provide this functionality.
> >>
> >> I am looking to use Bria which I believe is the gold standard and I
> sort of have push notification working on it. The problem is that Push
> Notification work best with multiple registration which is not supported
> with chan_sip.
> >>
> >> So I have two options I believe:
> >> • Migrate to PJSIP
> >> • Add Kamailio to Astlinux to sip in front of Asterisk
> >>
> >> Now 1) will certainly be the easiest however 2) is something that I
> would love to do as its going to be more secure with external connecting
> clients.
> >>
> >> Any comments would be greatly appreciated?
> >>
> >> Regards
> >> Michael Knill
> >
> > Hi Michael,
> >
> > I have similar thoughts and would be interested in a more universal
> solution as well.
> >
> > I have one big customer, where I use WireGuard plus mostly Linphone
> without Push Notifications.
> > And the customer is quite happy with it, although they are not using it
> excessively.
> >
> > I created a (combined) template for an IP-phone plus a softphone under
> the same extension, same CLID, but with different SIP account names (with
> chan_sip).
> >
> > I generated the dialplan (in the template as well) so that both phones
> ring parallel.
> >
> > Most depends how the mobile OS handles priority for the SIP client on
> the phone and the the sleep functions.
> >
> > For Push Notifications I found this:
> >
> > https://www.zoiper.com/en/tutorials/push-notifications
> > https://github.com/balusreekanth/ios-asterisk-push
> >
> > BTW: For simple notifications (only for my own phone) I use Pushover
> (works also via email) or prepaid SMS via Voipbuster.
> >
> > https://pushover.net
> > https://www.voipbuster.com/sms_rates
> >
> > Michael
>
> Has anyone tested "Acrobits Groundwire" or "Acrobits Softphone" ?
>
> It looks like many "custom" mobile UC apps are based on the Acrobits SDK.
>
> Lonnie
>
>
>
>
>
> _______________________________________________
> Astlinux-devel mailing list
> Ast...@li...
> https://lists.sourceforge.net/lists/listinfo/astlinux-devel
>
> _______________________________________________
> Astlinux-devel mailing list
> Ast...@li...
> https://lists.sourceforge.net/lists/listinfo/astlinux-devel
>
|
|
From: David K. <da...@ke...> - 2020-11-28 22:51:22
|
Looking at https://developer.apple.com/documentation/networkextension/personal_vpn/vpn_on_demand_rules it seems that Wireguard have implemented very basic VPN on-demand (turn it on when on cellular or when on wifi). What would be needed is much more granular control, at a minimum turn on for specific applications only. Even better would be connection specific rules. That feels like a lot of work which would have to get implemented in the Wireguard iOS app. David On Sat, Nov 28, 2020 at 5:35 PM David Kerr <da...@ke...> wrote: > I used to have 5061 open to anything, but I started to get attempts to > login and make calls, so I now restrict it to the Acrobits servers. For > the iPhone itself I have to be connected to my home network either directly > to the WiFi or by VPN (I use Wireguard). Acrobits (afaik) has no built-in > VPN. I have not tried the Wireguard on-demand setting but it is possible > that might work. > > I have not tried roaming between WiFi and Mobile. > > David > > On Sat, Nov 28, 2020 at 3:44 PM Michael Knill < > mic...@ip...> wrote: > >> Hi All >> >> >> >> Thanks for responding. >> >> Bria is certainly the gold standard however I have been down the Acrobits >> path and had a full demo of what they can provide. They have a full white >> labelling solution and there provisioning system is pretty cool. Just point >> your phone to a QR code. >> >> >> >> David are you only opening up 5061 to the push server? If so then your >> iPhone must never directly register to your system externally? >> >> Also another question; can you roam between Wi-Fi and 4G while on a call? >> >> >> >> Regards >> >> Michael Knill >> >> >> >> *From: *David Kerr <da...@ke...> >> *Reply to: *AstLinux Developers Mailing List < >> ast...@li...> >> *Date: *Sunday, 29 November 2020 at 2:48 am >> *To: *AstLinux Developers Mailing List < >> ast...@li...> >> *Subject: *Re: [Astlinux-devel] Recommendations for my architecture >> moving forward >> >> >> >> I have had Acrobits Softphone installed on my iPhone for several years. >> I don't use it a whole lot but it works well. It has push notification for >> when the softphone is "in the background." The way that works is that they >> have their own server that registers to your SIP port and when a call comes >> in their server generates a push to your iPhone and on receipt of that you >> open up the Softphone and somehow a handoff is done from their server to >> your phone. Just recently I contacted their tech support to find out what >> IP addresses I should use for their servers (they have multiple) so that I >> didn't have to open up 5061 to everyone. This is what I have in my custom >> rules... >> >> >> >> open_for_host_port() >> >> { >> >> local host="$1" >> >> local port="$2" >> >> local type="$3" >> >> local IFS=$'\n' >> >> local IPV4="$(host -t A $host)" >> >> local IPV6="$(host -t AAAA $host)" >> >> local IP="" >> >> >> >> for IP in $IPV4; do >> >> IP="${IP##* }" >> >> echo "[CUSTOM RULE] open port $port/$type for $host at $IP" >> >> ip4tables -A EXT_INPUT_CHAIN -s $IP -p $type -m $type --dport $port >> -j ACCEPT >> >> done >> >> >> >> for IP in $IPV6; do >> >> IP="${IP##* }" >> >> echo "[CUSTOM RULE] open port $port/$type for $host at $IP" >> >> ip6tables -A EXT_INPUT_CHAIN -s $IP -p $type -m $type --dport $port >> -j ACCEPT >> >> done >> >> } >> >> open_for_host_port "all.sipis.acrobits.cz" "5061" "tcp" >> >> >> >> >> >> >> >> On Sat, Nov 28, 2020 at 10:17 AM Lonnie Abelbeck < >> li...@lo...> wrote: >> >> >> >> > On Nov 28, 2020, at 5:17 AM, Michael Keuter <li...@mk...> >> wrote: >> > >> >> >> >> Am 27.11.2020 um 22:46 schrieb Michael Knill < >> mic...@ip...>: >> >> >> >> Hi Devs >> >> >> >> Hoping that I can get some advice from you in determining my >> development efforts moving forward. >> >> As driven by Covid, I have an increased number of customers that want >> mobile softphones and I have lost at least one customer and likely more >> because I have not been able to provide this functionality. >> >> >> >> I am looking to use Bria which I believe is the gold standard and I >> sort of have push notification working on it. The problem is that Push >> Notification work best with multiple registration which is not supported >> with chan_sip. >> >> >> >> So I have two options I believe: >> >> • Migrate to PJSIP >> >> • Add Kamailio to Astlinux to sip in front of Asterisk >> >> >> >> Now 1) will certainly be the easiest however 2) is something that I >> would love to do as its going to be more secure with external connecting >> clients. >> >> >> >> Any comments would be greatly appreciated? >> >> >> >> Regards >> >> Michael Knill >> > >> > Hi Michael, >> > >> > I have similar thoughts and would be interested in a more universal >> solution as well. >> > >> > I have one big customer, where I use WireGuard plus mostly Linphone >> without Push Notifications. >> > And the customer is quite happy with it, although they are not using it >> excessively. >> > >> > I created a (combined) template for an IP-phone plus a softphone under >> the same extension, same CLID, but with different SIP account names (with >> chan_sip). >> > >> > I generated the dialplan (in the template as well) so that both phones >> ring parallel. >> > >> > Most depends how the mobile OS handles priority for the SIP client on >> the phone and the the sleep functions. >> > >> > For Push Notifications I found this: >> > >> > https://www.zoiper.com/en/tutorials/push-notifications >> > https://github.com/balusreekanth/ios-asterisk-push >> > >> > BTW: For simple notifications (only for my own phone) I use Pushover >> (works also via email) or prepaid SMS via Voipbuster. >> > >> > https://pushover.net >> > https://www.voipbuster.com/sms_rates >> > >> > Michael >> >> Has anyone tested "Acrobits Groundwire" or "Acrobits Softphone" ? >> >> It looks like many "custom" mobile UC apps are based on the Acrobits SDK. >> >> Lonnie >> >> >> >> >> >> _______________________________________________ >> Astlinux-devel mailing list >> Ast...@li... >> https://lists.sourceforge.net/lists/listinfo/astlinux-devel >> >> _______________________________________________ >> Astlinux-devel mailing list >> Ast...@li... >> https://lists.sourceforge.net/lists/listinfo/astlinux-devel >> > |
|
From: Michael K. <mic...@ip...> - 2020-11-28 20:43:49
|
Hi All
Thanks for responding.
Bria is certainly the gold standard however I have been down the Acrobits path and had a full demo of what they can provide. They have a full white labelling solution and there provisioning system is pretty cool. Just point your phone to a QR code.
David are you only opening up 5061 to the push server? If so then your iPhone must never directly register to your system externally?
Also another question; can you roam between Wi-Fi and 4G while on a call?
Regards
Michael Knill
From: David Kerr <da...@ke...>
Reply to: AstLinux Developers Mailing List <ast...@li...>
Date: Sunday, 29 November 2020 at 2:48 am
To: AstLinux Developers Mailing List <ast...@li...>
Subject: Re: [Astlinux-devel] Recommendations for my architecture moving forward
I have had Acrobits Softphone installed on my iPhone for several years. I don't use it a whole lot but it works well. It has push notification for when the softphone is "in the background." The way that works is that they have their own server that registers to your SIP port and when a call comes in their server generates a push to your iPhone and on receipt of that you open up the Softphone and somehow a handoff is done from their server to your phone. Just recently I contacted their tech support to find out what IP addresses I should use for their servers (they have multiple) so that I didn't have to open up 5061 to everyone. This is what I have in my custom rules...
open_for_host_port()
{
local host="$1"
local port="$2"
local type="$3"
local IFS=$'\n'
local IPV4="$(host -t A $host)"
local IPV6="$(host -t AAAA $host)"
local IP=""
for IP in $IPV4; do
IP="${IP##* }"
echo "[CUSTOM RULE] open port $port/$type for $host at $IP"
ip4tables -A EXT_INPUT_CHAIN -s $IP -p $type -m $type --dport $port -j ACCEPT
done
for IP in $IPV6; do
IP="${IP##* }"
echo "[CUSTOM RULE] open port $port/$type for $host at $IP"
ip6tables -A EXT_INPUT_CHAIN -s $IP -p $type -m $type --dport $port -j ACCEPT
done
}
open_for_host_port "all.sipis.acrobits.cz<http://all.sipis.acrobits.cz>" "5061" "tcp"
On Sat, Nov 28, 2020 at 10:17 AM Lonnie Abelbeck <li...@lo...<mailto:li...@lo...>> wrote:
> On Nov 28, 2020, at 5:17 AM, Michael Keuter <li...@mk...<mailto:li...@mk...>> wrote:
>
>>
>> Am 27.11.2020 um 22:46 schrieb Michael Knill <mic...@ip...<mailto:mic...@ip...>>:
>>
>> Hi Devs
>>
>> Hoping that I can get some advice from you in determining my development efforts moving forward.
>> As driven by Covid, I have an increased number of customers that want mobile softphones and I have lost at least one customer and likely more because I have not been able to provide this functionality.
>>
>> I am looking to use Bria which I believe is the gold standard and I sort of have push notification working on it. The problem is that Push Notification work best with multiple registration which is not supported with chan_sip.
>>
>> So I have two options I believe:
>> • Migrate to PJSIP
>> • Add Kamailio to Astlinux to sip in front of Asterisk
>>
>> Now 1) will certainly be the easiest however 2) is something that I would love to do as its going to be more secure with external connecting clients.
>>
>> Any comments would be greatly appreciated?
>>
>> Regards
>> Michael Knill
>
> Hi Michael,
>
> I have similar thoughts and would be interested in a more universal solution as well.
>
> I have one big customer, where I use WireGuard plus mostly Linphone without Push Notifications.
> And the customer is quite happy with it, although they are not using it excessively.
>
> I created a (combined) template for an IP-phone plus a softphone under the same extension, same CLID, but with different SIP account names (with chan_sip).
>
> I generated the dialplan (in the template as well) so that both phones ring parallel.
>
> Most depends how the mobile OS handles priority for the SIP client on the phone and the the sleep functions.
>
> For Push Notifications I found this:
>
> https://www.zoiper.com/en/tutorials/push-notifications
> https://github.com/balusreekanth/ios-asterisk-push
>
> BTW: For simple notifications (only for my own phone) I use Pushover (works also via email) or prepaid SMS via Voipbuster.
>
> https://pushover.net
> https://www.voipbuster.com/sms_rates
>
> Michael
Has anyone tested "Acrobits Groundwire" or "Acrobits Softphone" ?
It looks like many "custom" mobile UC apps are based on the Acrobits SDK.
Lonnie
_______________________________________________
Astlinux-devel mailing list
Ast...@li...<mailto:Ast...@li...>
https://lists.sourceforge.net/lists/listinfo/astlinux-devel
|
|
From: David K. <da...@ke...> - 2020-11-28 15:48:50
|
I have had Acrobits Softphone installed on my iPhone for several years. I
don't use it a whole lot but it works well. It has push notification for
when the softphone is "in the background." The way that works is that they
have their own server that registers to your SIP port and when a call comes
in their server generates a push to your iPhone and on receipt of that you
open up the Softphone and somehow a handoff is done from their server to
your phone. Just recently I contacted their tech support to find out what
IP addresses I should use for their servers (they have multiple) so that I
didn't have to open up 5061 to everyone. This is what I have in my custom
rules...
open_for_host_port()
{
local host="$1"
local port="$2"
local type="$3"
local IFS=$'\n'
local IPV4="$(host -t A $host)"
local IPV6="$(host -t AAAA $host)"
local IP=""
for IP in $IPV4; do
IP="${IP##* }"
echo "[CUSTOM RULE] open port $port/$type for $host at $IP"
ip4tables -A EXT_INPUT_CHAIN -s $IP -p $type -m $type --dport $port -j
ACCEPT
done
for IP in $IPV6; do
IP="${IP##* }"
echo "[CUSTOM RULE] open port $port/$type for $host at $IP"
ip6tables -A EXT_INPUT_CHAIN -s $IP -p $type -m $type --dport $port -j
ACCEPT
done
}
open_for_host_port "all.sipis.acrobits.cz" "5061" "tcp"
On Sat, Nov 28, 2020 at 10:17 AM Lonnie Abelbeck <li...@lo...>
wrote:
>
>
> > On Nov 28, 2020, at 5:17 AM, Michael Keuter <li...@mk...>
> wrote:
> >
> >>
> >> Am 27.11.2020 um 22:46 schrieb Michael Knill <
> mic...@ip...>:
> >>
> >> Hi Devs
> >>
> >> Hoping that I can get some advice from you in determining my
> development efforts moving forward.
> >> As driven by Covid, I have an increased number of customers that want
> mobile softphones and I have lost at least one customer and likely more
> because I have not been able to provide this functionality.
> >>
> >> I am looking to use Bria which I believe is the gold standard and I
> sort of have push notification working on it. The problem is that Push
> Notification work best with multiple registration which is not supported
> with chan_sip.
> >>
> >> So I have two options I believe:
> >> • Migrate to PJSIP
> >> • Add Kamailio to Astlinux to sip in front of Asterisk
> >>
> >> Now 1) will certainly be the easiest however 2) is something that I
> would love to do as its going to be more secure with external connecting
> clients.
> >>
> >> Any comments would be greatly appreciated?
> >>
> >> Regards
> >> Michael Knill
> >
> > Hi Michael,
> >
> > I have similar thoughts and would be interested in a more universal
> solution as well.
> >
> > I have one big customer, where I use WireGuard plus mostly Linphone
> without Push Notifications.
> > And the customer is quite happy with it, although they are not using it
> excessively.
> >
> > I created a (combined) template for an IP-phone plus a softphone under
> the same extension, same CLID, but with different SIP account names (with
> chan_sip).
> >
> > I generated the dialplan (in the template as well) so that both phones
> ring parallel.
> >
> > Most depends how the mobile OS handles priority for the SIP client on
> the phone and the the sleep functions.
> >
> > For Push Notifications I found this:
> >
> > https://www.zoiper.com/en/tutorials/push-notifications
> > https://github.com/balusreekanth/ios-asterisk-push
> >
> > BTW: For simple notifications (only for my own phone) I use Pushover
> (works also via email) or prepaid SMS via Voipbuster.
> >
> > https://pushover.net
> > https://www.voipbuster.com/sms_rates
> >
> > Michael
>
> Has anyone tested "Acrobits Groundwire" or "Acrobits Softphone" ?
>
> It looks like many "custom" mobile UC apps are based on the Acrobits SDK.
>
> Lonnie
>
>
>
>
>
> _______________________________________________
> Astlinux-devel mailing list
> Ast...@li...
> https://lists.sourceforge.net/lists/listinfo/astlinux-devel
>
|
|
From: Lonnie A. <li...@lo...> - 2020-11-28 15:17:33
|
> On Nov 28, 2020, at 5:17 AM, Michael Keuter <li...@mk...> wrote: > >> >> Am 27.11.2020 um 22:46 schrieb Michael Knill <mic...@ip...>: >> >> Hi Devs >> >> Hoping that I can get some advice from you in determining my development efforts moving forward. >> As driven by Covid, I have an increased number of customers that want mobile softphones and I have lost at least one customer and likely more because I have not been able to provide this functionality. >> >> I am looking to use Bria which I believe is the gold standard and I sort of have push notification working on it. The problem is that Push Notification work best with multiple registration which is not supported with chan_sip. >> >> So I have two options I believe: >> • Migrate to PJSIP >> • Add Kamailio to Astlinux to sip in front of Asterisk >> >> Now 1) will certainly be the easiest however 2) is something that I would love to do as its going to be more secure with external connecting clients. >> >> Any comments would be greatly appreciated? >> >> Regards >> Michael Knill > > Hi Michael, > > I have similar thoughts and would be interested in a more universal solution as well. > > I have one big customer, where I use WireGuard plus mostly Linphone without Push Notifications. > And the customer is quite happy with it, although they are not using it excessively. > > I created a (combined) template for an IP-phone plus a softphone under the same extension, same CLID, but with different SIP account names (with chan_sip). > > I generated the dialplan (in the template as well) so that both phones ring parallel. > > Most depends how the mobile OS handles priority for the SIP client on the phone and the the sleep functions. > > For Push Notifications I found this: > > https://www.zoiper.com/en/tutorials/push-notifications > https://github.com/balusreekanth/ios-asterisk-push > > BTW: For simple notifications (only for my own phone) I use Pushover (works also via email) or prepaid SMS via Voipbuster. > > https://pushover.net > https://www.voipbuster.com/sms_rates > > Michael Has anyone tested "Acrobits Groundwire" or "Acrobits Softphone" ? It looks like many "custom" mobile UC apps are based on the Acrobits SDK. Lonnie |
28 messages has been excluded from this view by a project administrator.
