Menu
▾
▴
bproc-users — General discussion about BProc.
You can subscribe to this list here.
| 2001 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
(25) |
Nov
|
Dec
(22) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2002 |
Jan
(13) |
Feb
(22) |
Mar
(39) |
Apr
(10) |
May
(26) |
Jun
(23) |
Jul
(38) |
Aug
(20) |
Sep
(27) |
Oct
(76) |
Nov
(32) |
Dec
(11) |
| 2003 |
Jan
(8) |
Feb
(23) |
Mar
(12) |
Apr
(39) |
May
(1) |
Jun
(48) |
Jul
(35) |
Aug
(15) |
Sep
(60) |
Oct
(27) |
Nov
(9) |
Dec
(32) |
| 2004 |
Jan
(8) |
Feb
(16) |
Mar
(40) |
Apr
(25) |
May
(12) |
Jun
(33) |
Jul
(49) |
Aug
(39) |
Sep
(26) |
Oct
(47) |
Nov
(26) |
Dec
(36) |
| 2005 |
Jan
(29) |
Feb
(15) |
Mar
(22) |
Apr
(1) |
May
(8) |
Jun
(32) |
Jul
(11) |
Aug
(17) |
Sep
(9) |
Oct
(7) |
Nov
(15) |
Dec
|
|
From: Sadanand K. <sa...@ci...> - 2002-07-16 07:56:24
|
Our scheduling techniques involves selection of existing process' pids based on certain criterion and move(migrate) that process to specified node (also identified by our scheduling algorithm). Do you think this is possible using BPROC ? Hope you undertand my question. Please let me know if there is any confusion. ( As I understand bproc_execmove cannot be used in this situation). Thanks a lot for your timely help, Sadanand On Tue, 16 Jul 2002, Wilton Wong wrote: > > On Tue, 16 Jul 2002, Sadanand Kota wrote: > > > regarding clustermatic installation - My friend succeded in running > > bpmaster and bpslave. Actually we are trying our own scheduling algorithms > > using underlying process migration techniques from Mosix,Bproc. Is there a > > way to migrate process using BPROC through a program by calling some > > function and passing the required pid? ( bproc_move(node) migrates only > > the current process .. right? ) > > I'm not sure what you mean by "migrate" but you can also use bproc_execmove() > which will execve the program then move it to a specified node, or you could > also use bproc_rfork() and create an identical child program on a specified > node (same like fork) > > - Wilton > > > ----[ Wilton William Wong ]--------------------------------------------- > 11060-166 Avenue Ph : 01-780-456-9771 High Performance UNIX > Edmonton, Alberta FAX: 01-780-456-9772 and Linux Solutions > T5X 1Y3, Canada URL: http://www.harddata.com > -------------------------------------------------------[ Hard Data Ltd. ]---- > > > > ------------------------------------------------------- > This sf.net email is sponsored by: Jabber - The world's fastest growing > real-time communications platform! Don't just IM. Build it in! > http://www.jabber.com/osdn/xim > _______________________________________________ > BProc-users mailing list > BPr...@li... > https://lists.sourceforge.net/lists/listinfo/bproc-users > |
|
From: Wilton W. <ww...@ha...> - 2002-07-16 07:40:47
|
On Tue, 16 Jul 2002, Sadanand Kota wrote: > regarding clustermatic installation - My friend succeded in running > bpmaster and bpslave. Actually we are trying our own scheduling algorithms > using underlying process migration techniques from Mosix,Bproc. Is there a > way to migrate process using BPROC through a program by calling some > function and passing the required pid? ( bproc_move(node) migrates only > the current process .. right? ) I'm not sure what you mean by "migrate" but you can also use bproc_execmove() which will execve the program then move it to a specified node, or you could also use bproc_rfork() and create an identical child program on a specified node (same like fork) - Wilton ----[ Wilton William Wong ]--------------------------------------------- 11060-166 Avenue Ph : 01-780-456-9771 High Performance UNIX Edmonton, Alberta FAX: 01-780-456-9772 and Linux Solutions T5X 1Y3, Canada URL: http://www.harddata.com -------------------------------------------------------[ Hard Data Ltd. ]---- |
|
From: Sadanand K. <sa...@ci...> - 2002-07-16 07:09:22
|
The output of cat /proc/ksysm | grep fput is c01363a0 fput_Rsmp_18820440 Sada On Tue, 16 Jul 2002, Wilton Wong wrote: > > On Tue, 16 Jul 2002, Sadanand Kota wrote: > > > Ya, I am sure that my configured kernel source is in usr/src/linux > > (/usr/src/linux-2.4.18 to be precise and /usr/src/linux is a soft link to > > /usr/src/linux-2.4.18) > > Hmm.. interesting.. can you send me the output of > > cat /proc/ksyms | grep fput > > > With Mac address, how do I set a single machine as both amster and > > slave ( for testing purposes)? > > I don't belive that this is possible. > > - Wilton > > ----[ Wilton William Wong ]--------------------------------------------- > 11060-166 Avenue Ph : 01-780-456-9771 High Performance UNIX > Edmonton, Alberta FAX: 01-780-456-9772 and Linux Solutions > T5X 1Y3, Canada URL: http://www.harddata.com > -------------------------------------------------------[ Hard Data Ltd. ]---- > |
|
From: Wilton W. <ww...@ha...> - 2002-07-16 07:05:15
|
On Tue, 16 Jul 2002, Sadanand Kota wrote: > Ya, I am sure that my configured kernel source is in usr/src/linux > (/usr/src/linux-2.4.18 to be precise and /usr/src/linux is a soft link to > /usr/src/linux-2.4.18) Hmm.. interesting.. can you send me the output of cat /proc/ksyms | grep fput > With Mac address, how do I set a single machine as both amster and > slave ( for testing purposes)? I don't belive that this is possible. - Wilton ----[ Wilton William Wong ]--------------------------------------------- 11060-166 Avenue Ph : 01-780-456-9771 High Performance UNIX Edmonton, Alberta FAX: 01-780-456-9772 and Linux Solutions T5X 1Y3, Canada URL: http://www.harddata.com -------------------------------------------------------[ Hard Data Ltd. ]---- |
|
From: Sadanand K. <sa...@ci...> - 2002-07-16 06:56:00
|
Ya, I am sure that my configured kernel source is in usr/src/linux (/usr/src/linux-2.4.18 to be precise and /usr/src/linux is a soft link to /usr/src/linux-2.4.18) With Mac address, how do I set a single machine as both amster and slave ( for testing purposes)? Sorry If I am bugging you lot of questions. Thanks a lot for your patient replies. Sadanand On Tue, 16 Jul 2002, Wilton Wong wrote: > > On Tue, 16 Jul 2002, Sadanand Kota wrote: > > > Thanks again Wilton. > > I had not tried 'make install'. I just did that and it completed with no > > errors. But now what? When I say insmod bproc, it now gives > > Using /lib/modules/2.4.18/bproc/bproc.o > > /lib/modules/2.4.18/bproc/bproc.o: unresolved symbol d_alloc_Rsmp_83cf43fa > > /lib/modules/2.4.18/bproc/bproc.o: unresolved symbol d_instantiate_Rsmp_440f05d0 > > /lib/modules/2.4.18/bproc/bproc.o: unresolved symbol bproc_hook_sys_wait4_3_hook_Rsmp_bc34da6c > > /lib/modules/2.4.18/bproc/bproc.o: unresolved symbol bproc_hook_send_sig_info_hook_Rsmp_18fccf39 > > /lib/modules/2.4.18/bproc/bproc.o: unresolved symbol fput_Rsmp_2e40bae6 > > /lib/modules/2.4.18/bproc/bproc.o: unresolved symbol is_orphaned_pgrp_Rsmp_d0f4e83b > > > > Also, if try to run bpmaster, it says > > './bpmaster: BPROC_SYS_VERSION: Function not implemented'. > > Hmm.. looks like modversions is getting confused, are you sure that your > currently configured kernel source was in /usr/src/linux when you did a > make/make install ? > > > My /etc/bproc.conf is as follows > > bind manager 2223 > > node 127.0.0.1 127.0.0.2 > > 127.0.0.1 cannot be used as nodes.. 127.0.0.x addresses cannot be used, they > are reserved for local traffic. > > node entries look like: > > node <MAC address of compute node> > > ie: > > node 00:E0:81:20:44:BB > > > On running the command /etc/rc.d/init.d/beowulf start it says > > Configuring network interface (eth0):Error: No netmask given for interface > > > > In /etc/beowulf/config file, netfask for eth0 is properly defined as > > follows > > interface eth0 255.255.255.0. > > should be: > > interface <ifname> <ipaddress> <netmask> > > ie: > > interface eth0 10.0.0.1 255.255.255.0 > > also make sure the "iprange" is within the scope of your netmask > > - Wilton > > ----[ Wilton William Wong ]--------------------------------------------- > 11060-166 Avenue Ph : 01-780-456-9771 High Performance UNIX > Edmonton, Alberta FAX: 01-780-456-9772 and Linux Solutions > T5X 1Y3, Canada URL: http://www.harddata.com > -------------------------------------------------------[ Hard Data Ltd. ]---- > |
|
From: Wilton W. <ww...@ha...> - 2002-07-16 06:48:59
|
On Tue, 16 Jul 2002, Sadanand Kota wrote: > Thanks again Wilton. > I had not tried 'make install'. I just did that and it completed with no > errors. But now what? When I say insmod bproc, it now gives > Using /lib/modules/2.4.18/bproc/bproc.o > /lib/modules/2.4.18/bproc/bproc.o: unresolved symbol d_alloc_Rsmp_83cf43fa > /lib/modules/2.4.18/bproc/bproc.o: unresolved symbol d_instantiate_Rsmp_440f05d0 > /lib/modules/2.4.18/bproc/bproc.o: unresolved symbol bproc_hook_sys_wait4_3_hook_Rsmp_bc34da6c > /lib/modules/2.4.18/bproc/bproc.o: unresolved symbol bproc_hook_send_sig_info_hook_Rsmp_18fccf39 > /lib/modules/2.4.18/bproc/bproc.o: unresolved symbol fput_Rsmp_2e40bae6 > /lib/modules/2.4.18/bproc/bproc.o: unresolved symbol is_orphaned_pgrp_Rsmp_d0f4e83b > > Also, if try to run bpmaster, it says > './bpmaster: BPROC_SYS_VERSION: Function not implemented'. Hmm.. looks like modversions is getting confused, are you sure that your currently configured kernel source was in /usr/src/linux when you did a make/make install ? > My /etc/bproc.conf is as follows > bind manager 2223 > node 127.0.0.1 127.0.0.2 127.0.0.1 cannot be used as nodes.. 127.0.0.x addresses cannot be used, they are reserved for local traffic. node entries look like: node <MAC address of compute node> ie: node 00:E0:81:20:44:BB > On running the command /etc/rc.d/init.d/beowulf start it says > Configuring network interface (eth0):Error: No netmask given for interface > > In /etc/beowulf/config file, netfask for eth0 is properly defined as > follows > interface eth0 255.255.255.0. should be: interface <ifname> <ipaddress> <netmask> ie: interface eth0 10.0.0.1 255.255.255.0 also make sure the "iprange" is within the scope of your netmask - Wilton ----[ Wilton William Wong ]--------------------------------------------- 11060-166 Avenue Ph : 01-780-456-9771 High Performance UNIX Edmonton, Alberta FAX: 01-780-456-9772 and Linux Solutions T5X 1Y3, Canada URL: http://www.harddata.com -------------------------------------------------------[ Hard Data Ltd. ]---- |
|
From: Sadanand K. <sa...@ci...> - 2002-07-16 06:37:50
|
Thanks again Wilton. I had not tried 'make install'. I just did that and it completed with no errors. But now what? When I say insmod bproc, it now gives Using /lib/modules/2.4.18/bproc/bproc.o /lib/modules/2.4.18/bproc/bproc.o: unresolved symbol d_alloc_Rsmp_83cf43fa /lib/modules/2.4.18/bproc/bproc.o: unresolved symbol d_instantiate_Rsmp_440f05d0 /lib/modules/2.4.18/bproc/bproc.o: unresolved symbol bproc_hook_sys_wait4_3_hook_Rsmp_bc34da6c /lib/modules/2.4.18/bproc/bproc.o: unresolved symbol bproc_hook_send_sig_info_hook_Rsmp_18fccf39 /lib/modules/2.4.18/bproc/bproc.o: unresolved symbol fput_Rsmp_2e40bae6 /lib/modules/2.4.18/bproc/bproc.o: unresolved symbol is_orphaned_pgrp_Rsmp_d0f4e83b Also, if try to run bpmaster, it says './bpmaster: BPROC_SYS_VERSION: Function not implemented'. My /etc/bproc.conf is as follows bind manager 2223 node 127.0.0.1 127.0.0.2 Theres no specific reason why I am trying custom compiled kernel. My partner( sitting right next to me now) is trying the clustematic binries (i.e he download all rpms from clustermatic.org). He also has a problem as follows - On running the command /etc/rc.d/init.d/beowulf start it says Configuring network interface (eth0):Error: No netmask given for interface In /etc/beowulf/config file, netfask for eth0 is properly defined as follows interface eth0 255.255.255.0. Also, after installation of clustermatic RPMS, upon booting the machine says, etho - unknown hosts. This problem was not there before installation of clustermatic RPM. Any idea for either(custom kernel or clustermatic RPM) are highly appreciated. Thanks, Sadanand --------------------------------------------------------------------------- On Tue, 16 Jul 2002, Wilton Wong wrote: > > On Tue, 16 Jul 2002, Sadanand Kota wrote: > > > Thanks a lot for your reply Mr.Wilton. I am attaching my .config file with > > this mail. > > Mr. Wong actually... but Call me Wilton.. (no one calls me Mr. Wong ;) > > > I did not understand what you meant by recomplie bproc/vmadump. I tried it > > now by saying 'make' inside vmadump directory. Its giving errors as > > Did you do a "make" inside the toplevel directory ? and then a "make install" > also please check that your kernel source is in /usr/src/linux. > > > insmod ksyscall. it says > > ksyscall.o: couldn't find the kernel version the module was compiled for > > That's correct ksyscall.o is not a kernel module, the two modules in the bproc > pacakge are called vmadump.o and bproc.o > > Is there a reason that you need a custom compiled kernel for your application ? > have you tried the clustermatic binaries ? > > - Wilton > > PS. looking at your kernel configuration you will need to recompile it with > ramdisk support if you wish this kernel to be used as the "normally running" > kernel on compute nodes. > > ----[ Wilton William Wong ]--------------------------------------------- > 11060-166 Avenue Ph : 01-780-456-9771 High Performance UNIX > Edmonton, Alberta FAX: 01-780-456-9772 and Linux Solutions > T5X 1Y3, Canada URL: http://www.harddata.com > -------------------------------------------------------[ Hard Data Ltd. ]---- > |
|
From: Wilton W. <ww...@ha...> - 2002-07-16 06:32:41
|
Did I say ramdisk.. cause I meant ROMFS.. but it looks like you already have it in there ;) sorry my mistake. - Wilton ----[ Wilton William Wong ]--------------------------------------------- 11060-166 Avenue Ph : 01-780-456-9771 High Performance UNIX Edmonton, Alberta FAX: 01-780-456-9772 and Linux Solutions T5X 1Y3, Canada URL: http://www.harddata.com -------------------------------------------------------[ Hard Data Ltd. ]---- |
|
From: Wilton W. <ww...@ha...> - 2002-07-16 06:17:29
|
On Tue, 16 Jul 2002, Sadanand Kota wrote: > Thanks a lot for your reply Mr.Wilton. I am attaching my .config file with > this mail. Mr. Wong actually... but Call me Wilton.. (no one calls me Mr. Wong ;) > I did not understand what you meant by recomplie bproc/vmadump. I tried it > now by saying 'make' inside vmadump directory. Its giving errors as Did you do a "make" inside the toplevel directory ? and then a "make install" also please check that your kernel source is in /usr/src/linux. > insmod ksyscall. it says > ksyscall.o: couldn't find the kernel version the module was compiled for That's correct ksyscall.o is not a kernel module, the two modules in the bproc pacakge are called vmadump.o and bproc.o Is there a reason that you need a custom compiled kernel for your application ? have you tried the clustermatic binaries ? - Wilton PS. looking at your kernel configuration you will need to recompile it with ramdisk support if you wish this kernel to be used as the "normally running" kernel on compute nodes. ----[ Wilton William Wong ]--------------------------------------------- 11060-166 Avenue Ph : 01-780-456-9771 High Performance UNIX Edmonton, Alberta FAX: 01-780-456-9772 and Linux Solutions T5X 1Y3, Canada URL: http://www.harddata.com -------------------------------------------------------[ Hard Data Ltd. ]---- |
|
From: Sadanand K. <sa...@ci...> - 2002-07-16 06:06:25
|
Thanks a lot for your reply Mr.Wilton. I am attaching my .config file with this mail. I did not understand what you meant by recomplie bproc/vmadump. I tried it now by saying 'make' inside vmadump directory. Its giving errors as In function `vmadump': /root/BPROC/bproc-3.1.10/vmadump/vmadtest.c:63: undefined reference to `bproc_dump' /tmp/ccJfzPlO.o: In function `vmaundump': /root/BPROC/bproc-3.1.10/vmadump/vmadtest.c:67: undefined reference to `bproc_undump' /tmp/ccJfzPlO.o: In function `vmaexecdump': /root/BPROC/bproc-3.1.10/vmadump/vmadtest.c:71: undefined reference to `bproc_execdump' collect2: ld returned 1 exit status make: *** [vmadtest] Error 1 On insmod bproc.o, the first few lines or errors are as follows insmod bproc.o bproc.o: unresolved symbol d_alloc_Rsmp_83cf43fa bproc.o: unresolved symbol d_instantiate_Rsmp_440f05d0 bproc.o: unresolved symbol bproc_hook_sys_wait4_3_hook_Rsmp_bc34da6c bproc.o: unresolved symbol bproc_hook_send_sig_info_hook_Rsmp_18fccf39 bproc.o: unresolved symbol fput_Rsmp_2e40bae6 bproc.o: unresolved symbol is_orphaned_pgrp_Rsmp_d0f4e83b bproc.o: unresolved symbol notify_parent_Rsmp_eec9348d bproc.o: unresolved symbol child_reaper_Rsmp_d5b049c2 bproc.o: unresolved symbol bproc_hook_do_fork_2_hook_Rsmp_4b7e9ea9 bproc.o: unresolved symbol bproc_hook_do_execve_hook_Rsmp_a867c871 bproc.o: unresolved symbol sockfs_dentry_operations_Rsmp_f984a945 On insmod ksyscall. it says ksyscall.o: couldn't find the kernel version the module was compiled for Any idea, Thanks, Sadanand On Mon, 15 Jul 2002, Wilton Wong wrote: > > On Tue, 16 Jul 2002, Sadanand Kota wrote: > > > Hi, > > I just downloaded bproc-3.1.10 from sourceforage. After I compile new > > kernel, I succeded in 'make'. But I am unable to install through insmod or > > modprobe. Its giving lots of unresoled symbol errors as soon as I say > > insmod ksyscall.o or insmod vmadump.o. I read one of your messages on > > mailing list saying not to do insmod, instead to try modprobe bproc. > > When I say modprobe bproc, it says 'cant locate the module bproc'. I again > > get lots of unresolved symbol errors on insmod bproc.o. > > > > Can someone help in this regard, > > > > Thanks, > > Sadanand > > > > PS: I am haing linux-2.4.18 and I also applied the patch which comes with > > bproc for this kernel. > > What are the unresolved symbols tht you are getting ? and did you recompile > bproc/vmadump from bproc-3.1.10 sources ? also can you send me your .config > file ? > > - Wilton > > ----[ Wilton William Wong ]--------------------------------------------- > 11060-166 Avenue Ph : 01-780-456-9771 High Performance UNIX > Edmonton, Alberta FAX: 01-780-456-9772 and Linux Solutions > T5X 1Y3, Canada URL: http://www.harddata.com > -------------------------------------------------------[ Hard Data Ltd. ]---- > > > > ------------------------------------------------------- > This sf.net email is sponsored by: Jabber - The world's fastest growing > real-time communications platform! Don't just IM. Build it in! > http://www.jabber.com/osdn/xim > _______________________________________________ > BProc-users mailing list > BPr...@li... > https://lists.sourceforge.net/lists/listinfo/bproc-users > |
|
From: Wilton W. <ww...@ha...> - 2002-07-16 05:47:15
|
On Tue, 16 Jul 2002, Sadanand Kota wrote: > Hi, > I just downloaded bproc-3.1.10 from sourceforage. After I compile new > kernel, I succeded in 'make'. But I am unable to install through insmod or > modprobe. Its giving lots of unresoled symbol errors as soon as I say > insmod ksyscall.o or insmod vmadump.o. I read one of your messages on > mailing list saying not to do insmod, instead to try modprobe bproc. > When I say modprobe bproc, it says 'cant locate the module bproc'. I again > get lots of unresolved symbol errors on insmod bproc.o. > > Can someone help in this regard, > > Thanks, > Sadanand > > PS: I am haing linux-2.4.18 and I also applied the patch which comes with > bproc for this kernel. What are the unresolved symbols tht you are getting ? and did you recompile bproc/vmadump from bproc-3.1.10 sources ? also can you send me your .config file ? - Wilton ----[ Wilton William Wong ]--------------------------------------------- 11060-166 Avenue Ph : 01-780-456-9771 High Performance UNIX Edmonton, Alberta FAX: 01-780-456-9772 and Linux Solutions T5X 1Y3, Canada URL: http://www.harddata.com -------------------------------------------------------[ Hard Data Ltd. ]---- |
|
From: Sadanand K. <sa...@ci...> - 2002-07-16 05:39:01
|
Hi, I just downloaded bproc-3.1.10 from sourceforage. After I compile new kernel, I succeded in 'make'. But I am unable to install through insmod or modprobe. Its giving lots of unresoled symbol errors as soon as I say insmod ksyscall.o or insmod vmadump.o. I read one of your messages on mailing list saying not to do insmod, instead to try modprobe bproc. When I say modprobe bproc, it says 'cant locate the module bproc'. I again get lots of unresolved symbol errors on insmod bproc.o. Can someone help in this regard, Thanks, Sadanand PS: I am haing linux-2.4.18 and I also applied the patch which comes with bproc for this kernel. |
|
From: Erik A. H. <er...@he...> - 2002-07-15 14:51:13
|
On Thu, Jul 11, 2002 at 03:50:05PM -0700, Larry Baker wrote: > Is line continuation allowed in /etc/beowulf/config? If so, what is the syntax? (E.g., "\<newline>" as in C?) No. > Are multiple "libraries" lines legal? Do they all get concatenated into a single list? How long (no. of characters) can the list of libraries be? Yes. I don't believe there's any length limit on the lines. You can also specify directory names instead of libraries and use wildcards. - Erik -- Erik Arjan Hendriks Printed On 100 Percent Recycled Electrons er...@he... Contents may settle during shipment |
|
From: Larry B. <ba...@us...> - 2002-07-12 22:14:34
|
I have a testbed Clustermatic cluster running on a 1.6AGHz P4 master = with 2 diskless 1.8AGhz P4 slaves. I NFS mount /bin, /opt, /sbin, and = /usr read-only; /var is mounted read-write to a slave-specific = subdirectory; /home is mounted read-write and is shared across all = nodes. I've installed the Intel Fortran and C/C++ compilers, and their = Math and Performance libraries. I have compiled and tested the MPICH = libraries from ANL for parallel processing. The only tests in the MPICH = test suite that fail are the IO tests. The MPICH manual warns about sharing NFS filesystems for parallel = read-write access. I followed their recommendation to use the "noac" = option. However, the MPICH IO test suite still fails. I suspect they = are trying to use features of NFS that are not available when the = "nolock" option is used. (The test log says something about = read-modify-write.) However, the slave nodes do not boot correctly if I = remove the "nolock" option from any of the NFS mounts in = /etc/beowulf/fstab. This seems to be because neither portmapper or = lockd are running on the slave nodes. I can run portmapper (bpsh 0 = initlog -c portmap), and bpsh 0 rpcinfo -p then shows the 111 portmapper = udp/tcp service. But, when I try to run lockd (bpsh 0 initlog -c = rpc.lockd), it fails with "lockdsvc: Function not implemented". I tried = to copy /etc/rpc (bpcp /etc/rpc to 0:/etc) before starting portmapper, = but that did not help. Assuming I can get portmapper and lockd running, I envision a = modification to /usr/lib/beoboot/bin/setup_fs that would check an NFS = mount for the "nolock" option, and would start up portmapper and lockd = if necessary before doing a mount that did not have the "nolock" option. = Does this sound reasonable? Can anyone help from here? Thanks, Larry Baker |
|
From: Larry B. <ba...@us...> - 2002-07-11 22:43:27
|
Is line continuation allowed in /etc/beowulf/config? If so, what is the = syntax? (E.g., "\<newline>" as in C?) Are multiple "libraries" lines legal? Do they all get concatenated into = a single list? How long (no. of characters) can the list of libraries = be? |
|
From: Niels W. <Nie...@um...> - 2002-07-09 10:50:13
|
I have finally been able to sucessfully install the network drivers, after much hastle. The easiest solution was of course to compile 3c59x and pci-scan outside the kernel, and boot them in the right module sections. The problem arrises that pci-scan doen't get inserted in the modules of stage 1 and stage 2 kernels due to a very "simple" line in beoboot, which changes a - to a _ ! Renaming pci-scan to pci_scan solves that problem, and my nodes are now humming nicely. This shows that the 3c59x from the "default" clustermatic/bproc 2.4.18 kernel is indeed buggy for certain (if not all) cards! Niels -- Dr Niels R. Walet http://www.phy.umist.ac.uk/Theory/people/walet.html Dept. of Physics, UMIST, P.O. Box 88, Manchester, M60 1QD, U.K. Phone: +44(0)161-2003693 Fax: +44(0)161-2004303 Nie...@um... |
|
From: Niels W. <Nie...@um...> - 2002-07-08 10:59:12
|
I have been having some severe problems after having convertes to
clustermatic (from scyld).
It seems that the 3c59x module included in the lanl kernels is not
compatible with 2 kernel monte (I get autonegotation problems on my
switch(es), whatever the settings on my 3c980, they will always fall
back to half duplex/10MBits). I tried to copy Don Becker's latest
driver, and rebuild the kernels from the sources, using the spec file
from the clustermatic rpm. The issue is that this only produces an .i386
set of kernel files (having compiled for i386, i686 and athlon). My
knowledge of rpm specs is limited, I can see where compilation takes
place, but I can't see the final {i686,athlon}packages being build. Any
suggestions?
Furthermore, there is a complaint about missing symbols in the 3c59x.o
module on install of the i386 packages - and the interface hangs.
Does anyone of the list have any ideas?
--
Dr Niels R. Walet http://www.phy.umist.ac.uk/Theory/people/walet.html
Dept. of Physics, UMIST, P.O. Box 88, Manchester, M60 1QD, U.K.
Phone: +44(0)161-2003693 Fax: +44(0)161-2004303 Nie...@um...
|
|
From: Alexander V. <vi...@ma...> - 2002-07-06 20:56:13
|
On Sat, 6 Jul 2002, Albert D. Cahalan wrote:
> It made the 2.1.44 kernel horribly bad. It used to provide
> the POSIX-prohibited ability to do rmdir(".") safely.
"Safely" is not the word I'd use for something that allowed to
subvert sticky bit on directories, to start with...
|
|
From: Albert D. C. <aca...@cs...> - 2002-07-06 20:36:29
|
J.A. Magallon writes:
> I was looking info about ClusterNFS, and I found some info on 'transname',
> ie, do the same as cNFS but on the kernel server.
>
> Is that project still alive ? Is there any patch for 2.4 ?
The transname+omirr patch went in with kernel 2.1.43, but
only a few identifiable scraps remain in recent kernels.
That patch was of major importance to the history of Linux.
It is the reason why Linux has the dcache stuff, which lets
us have multiple mount points, per-process namespaces, pathnames
in /proc symlinks, and an efficient getcwd() system call.
It made the 2.1.44 kernel horribly bad. It used to provide
the POSIX-prohibited ability to do rmdir(".") safely.
You can see bits of the patch in the 2.4 kernel source:
--- fs/namei.c ---
/* [Feb-1997 T. Schoebel-Theuer]
* Fundamental changes in the pathname lookup mechanisms (namei)
* were necessary because of omirr. The reason is that omirr needs
* to know the _real_ pathname, not the user-supplied one, in case
* of symlinks (and also when transname replacements occur).
------------------
--- include/linux/dcache.h ---
/* appendix may either be NULL or be used for transname suffixes */
extern struct dentry * d_lookup(struct dentry *, struct qstr *);
------------------------------
--- Documentation/Configure.help ---
LocalWords: prio Micom xIO dwmw rimi OMIRR omirr omirrd unicode ntfs cmu NIC
------------------------------------
The actual transname stuff never made it into the kernel.
We just have some of the support code. Prior to the patch,
our filesystem code was like what BSD uses. Now we have
something which is more like UnixWare, but likely better.
I think Bill Hawes was responsible for this code after it
got in.
You might be able to make up for the missing functionality
with file-on-file mounts:
touch /tmp/mtab
mount --bind /tmp/mtab /etc/mtab
|
|
From: J.A. M. <jam...@ab...> - 2002-07-06 18:49:08
|
Hi all... I was looking info about ClusterNFS, and I found some info on 'transname', ie, do the same as cNFS but on the kernel server. Is that project still alive ? Is there any patch for 2.4 ? TIA -- J.A. Magallon \ Software is like sex: It's better when it's free mailto:jam...@ab... \ -- Linus Torvalds, FSF T-shirt Linux werewolf 2.4.19-rc1-jam1, Mandrake Linux 8.3 (Cooker) for i586 gcc (GCC) 3.1.1 (Mandrake Linux 8.3 3.1.1-0.7mdk) |
|
From: J.A. M. <jam...@ab...> - 2002-07-01 23:32:44
|
On 2002.07.02 Erik Arjan Hendriks wrote:
>
>Ok... Hunted this one down. The problem is with the linuxthreads
>library as far as I can tell. What they've done is provided a new
>fork function in place of libc's fork function. The pthreads fork
>does some magical stuff that bproc_rfork (obviously) doesn't do.
>
>You can create the same problem w/o bproc by replacing the bproc_rfork
>call with a call to __libc_fork(). That will bypass the pthreads
>magic too.
>
Someone in the kernel list suggested that, pthreads was overwriting something
But my tests show that what is getting overwritten is the 'clone()' call.
AFAIK, everything in based on the clone() system call: fork uses
clone() internally, pthreads are (quote) 'a simple POSIX threads library
implemented directly on top of clone()'.
And my test also breaks if I use clone():
void nslave()
{
...
for (i=0; i<nprocs; i++)
{
puts("about to clone...");
tid[i] = clone(pslave,
stack[i]+STSZ-1,CLONE_VM|SIGCHLD,
spawned);
puts("done");
spawned++;
}
...
}
I do not use any pthreads call. I just use clone. If I link with
-lpthread, it breaks. If not, it works. So I will look at clone.
A possible workaround could be to use NPTHREADS from IBM.
--
J.A. Magallon \ Software is like sex: It's better when it's free
mailto:jam...@ab... \ -- Linus Torvalds, FSF T-shirt
Linux werewolf 2.4.19-rc1-jam1, Mandrake Linux 8.3 (Cooker) for i586
gcc (GCC) 3.1.1 (Mandrake Linux 8.3 3.1.1-0.7mdk)
|
|
From: Erik A. H. <er...@he...> - 2002-07-01 22:02:30
|
On Sat, Jun 22, 2002 at 03:09:26PM +0200, J.A. Magallon wrote: > > On 2002.06.22 Erik Arjan Hendriks wrote: > >On Thu, Jun 20, 2002 at 12:29:20AM +0200, J.A. Magallon wrote: > >> Hi, bproc users and developers. > >> > >> I think I have found a weird bug...if it is so, and I am not misunderstanding > >> something. > > > >This is weird. Can you send me the snippet of test code so that I > >reproduce it? It seems like the sort of thing that should be easy to > >hunt down if I can reproduce it. > > > > Program and Makefile are attached. Defining DO_RFORK uses rfork(), > and without it does a fork()-move(). > There is a big sleep in pslave() to have time to see process structure > with pstree -p. > > Thanks, I hope you can get what is wrong (or confirm I'm on a _big_ > mistake). Ok... Hunted this one down. The problem is with the linuxthreads library as far as I can tell. What they've done is provided a new fork function in place of libc's fork function. The pthreads fork does some magical stuff that bproc_rfork (obviously) doesn't do. You can create the same problem w/o bproc by replacing the bproc_rfork call with a call to __libc_fork(). That will bypass the pthreads magic too. I don't see any reasonable way to work around this without modifying the pthreads library. All the datastructures and constructors are declared static. Pretty gross, IMO. - Erik -- Erik Arjan Hendriks Printed On 100 Percent Recycled Electrons er...@he... Contents may settle during shipment |
|
From: Aaron M. <aa...@cs...> - 2002-06-25 15:13:42
|
Well I upgradd to matched 2.2.4-35 glibc, recompiled and rebooted and now it seems to work, thanks for the helpful bits! Aaron -- _______________________________________________________ Aaron Macks(aa...@cs...) My sheep has seven gall bladders, that makes me the King of the Universe! |
|
From: Aaron M. <aa...@cs...> - 2002-06-25 15:02:14
|
bpslave segfaults before the connection is fully established(ie nothing like 'bpmaster: Setting status of node 0 to 4.'). Aaron > >Can you elaborate on the "the slave crashes"? Is that in the kernel? >Is there an oops? Does the daemon segfault or something? Are things >segfaulting when they migrate to the node? > >You will need to have the same versions of shared libraries for >process migration to work. A mismatch there should only cause >segfaults in processes trying to migrate. > >- Erik >-- >Erik Arjan Hendriks Printed On 100 Percent Recycled Electrons >er...@he... Contents may settle during shipment > -- _______________________________________________________ Aaron Macks(aa...@cs...) My sheep has seven gall bladders, that makes me the King of the Universe! |
|
From: Erik A. H. <er...@he...> - 2002-06-25 13:28:39
|
On Tue, Jun 25, 2002 at 01:41:17AM -0400, Aaron Macks wrote: > I'm in the process of configuring bproc on 5 Athlon 650's. Currently > I have one machine working properly(it can run master and a slave and > move processes) and one almost working. The problem is that when > starting the bpslave program, immediatly after the bpmaster(running > as -dv) shows a connection, the slave crashes. The only difference > between the systems is currently a 2 sub-version difference in > glibc(not currently running the patched version on either) do I need > to upgrade(required mosh of RH updated) glibc or is there something > else? Can you elaborate on the "the slave crashes"? Is that in the kernel? Is there an oops? Does the daemon segfault or something? Are things segfaulting when they migrate to the node? You will need to have the same versions of shared libraries for process migration to work. A mismatch there should only cause segfaults in processes trying to migrate. - Erik -- Erik Arjan Hendriks Printed On 100 Percent Recycled Electrons er...@he... Contents may settle during shipment |
1488 messages has been excluded from this view by a project administrator.
