Menu
▾
▴
treebase-devel — TreeBASE developer mailing list
You can subscribe to this list here.
| 2009 |
Jan
|
Feb
|
Mar
(1) |
Apr
(41) |
May
(41) |
Jun
(50) |
Jul
(14) |
Aug
(21) |
Sep
(37) |
Oct
(8) |
Nov
(4) |
Dec
(135) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2010 |
Jan
(145) |
Feb
(110) |
Mar
(216) |
Apr
(101) |
May
(42) |
Jun
(42) |
Jul
(23) |
Aug
(17) |
Sep
(33) |
Oct
(15) |
Nov
(18) |
Dec
(6) |
| 2011 |
Jan
(8) |
Feb
(10) |
Mar
(8) |
Apr
(41) |
May
(48) |
Jun
(62) |
Jul
(7) |
Aug
(9) |
Sep
(7) |
Oct
(11) |
Nov
(49) |
Dec
(1) |
| 2012 |
Jan
(17) |
Feb
(63) |
Mar
(4) |
Apr
(13) |
May
(17) |
Jun
(21) |
Jul
(10) |
Aug
(10) |
Sep
|
Oct
|
Nov
|
Dec
(16) |
| 2013 |
Jan
(10) |
Feb
|
Mar
(1) |
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2014 |
Jan
(5) |
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
(5) |
Nov
|
Dec
|
| 2015 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
(1) |
Nov
|
Dec
|
|
From: Mattison W. <mat...@ne...> - 2012-12-19 21:13:43
|
I can disable load balancing for download requests where the querystring includes x-access-code. Is that worth a try? The other option would be to load balance everything rather than just downloads. I can direct requests to the backend tomcat instances based on JSESSIONID. I think it would be fairly easy to isolate submissions from Dryad and make sure they always go to a primary tomcat instance if that is the only impediment. Regards, Mattison On Wed, Dec 19, 2012 at 9:40 AM, William Piel <wil...@ya...> wrote: > > Ah -- the downloads are working again… so indeed, the problem was with the > load balancing. > > bp > > > On Dec 19, 2012, at 8:36 AM, Mattison Ward wrote: > > I have disabled load balancing on treebase.org. I will do testing > with these x-access-code links on the dev system and work on the best > way to load balance them. > > Regards, > > Mattison > > On Tue, Dec 18, 2012 at 10:56 PM, William Piel <wil...@ya...> > wrote: > > > On Dec 18, 2012, at 8:49 PM, Shyket, Harry wrote: > > > I don’t think it may be the caching as it seems to redirect to the access > > violation page. Can we look at another study review to see if this is > > happening? > > > > Same behavior here: > > > http://purl.org/phylo/treebase/phylows/study/TB2:S9930?x-access-code=fd6c06635b71a5adcceacd8c6792a6c8&format=html > > > bp > > > |
|
From: William P. <wil...@ya...> - 2012-12-19 14:40:39
|
Ah -- the downloads are working again… so indeed, the problem was with the load balancing. bp On Dec 19, 2012, at 8:36 AM, Mattison Ward wrote: > I have disabled load balancing on treebase.org. I will do testing > with these x-access-code links on the dev system and work on the best > way to load balance them. > > Regards, > > Mattison > > On Tue, Dec 18, 2012 at 10:56 PM, William Piel <wil...@ya...> wrote: >> >> On Dec 18, 2012, at 8:49 PM, Shyket, Harry wrote: >> >> I don’t think it may be the caching as it seems to redirect to the access >> violation page. Can we look at another study review to see if this is >> happening? >> >> >> Same behavior here: >> >> http://purl.org/phylo/treebase/phylows/study/TB2:S9930?x-access-code=fd6c06635b71a5adcceacd8c6792a6c8&format=html >> >> bp > |
|
From: Shyket, H. <har...@ya...> - 2012-12-19 14:36:06
|
I just tried the previous example that Bill gave and it worked. Access Code http://purl.org/phylo/treebase/phylows/study/TB2:S13742?x-access-code=dfe493384bc32ede886fdcb0ba9bb17c&format=html File http://purl.org/phylo/treebase/phylows/study/TB2:S13742?format=nexus&x-access-code=dfe493384bc32ede886fdcb0ba9bb17c I think for the load balancing to work properly we need to consider moving the session data to the database. Thanks, Harry Shyket Digital Media Specialist Yale University Peabody Museum ph. 203-436-9428 har...@ya... -----Original Message----- From: mat...@gm... [mailto:mat...@gm...] On Behalf Of Mattison Ward Sent: Wednesday, December 19, 2012 8:37 AM To: William Piel Cc: Shyket, Harry; TreeBASE devel Subject: Re: Reviewer access broken I have disabled load balancing on treebase.org. I will do testing with these x-access-code links on the dev system and work on the best way to load balance them. Regards, Mattison On Tue, Dec 18, 2012 at 10:56 PM, William Piel <wil...@ya...> wrote: > > On Dec 18, 2012, at 8:49 PM, Shyket, Harry wrote: > > I don't think it may be the caching as it seems to redirect to the > access violation page. Can we look at another study review to see if > this is happening? > > > Same behavior here: > > http://purl.org/phylo/treebase/phylows/study/TB2:S9930?x-access-code=f > d6c06635b71a5adcceacd8c6792a6c8&format=html > > bp > > -- Mattison Ward NESCent at Duke University 2024 W. Main Street, Suite A200 Durham, NC 27705-4667 919-668-4585 (desk) 919-668-4551 (alternate) 919-668-9198 (fax) |
|
From: Mattison W. <mat...@ne...> - 2012-12-19 13:37:25
|
I have disabled load balancing on treebase.org. I will do testing with these x-access-code links on the dev system and work on the best way to load balance them. Regards, Mattison On Tue, Dec 18, 2012 at 10:56 PM, William Piel <wil...@ya...> wrote: > > On Dec 18, 2012, at 8:49 PM, Shyket, Harry wrote: > > I don’t think it may be the caching as it seems to redirect to the access > violation page. Can we look at another study review to see if this is > happening? > > > Same behavior here: > > http://purl.org/phylo/treebase/phylows/study/TB2:S9930?x-access-code=fd6c06635b71a5adcceacd8c6792a6c8&format=html > > bp > > -- Mattison Ward NESCent at Duke University 2024 W. Main Street, Suite A200 Durham, NC 27705-4667 919-668-4585 (desk) 919-668-4551 (alternate) 919-668-9198 (fax) |
|
From: William P. <wil...@ya...> - 2012-12-19 03:56:58
|
On Dec 18, 2012, at 8:49 PM, Shyket, Harry wrote: > I don’t think it may be the caching as it seems to redirect to the access violation page. Can we look at another study review to see if this is happening? Same behavior here: http://purl.org/phylo/treebase/phylows/study/TB2:S9930?x-access-code=fd6c06635b71a5adcceacd8c6792a6c8&format=html bp |
|
From: Shyket, H. <har...@ya...> - 2012-12-19 01:50:06
|
I don't think it may be the caching as it seems to redirect to the access violation page. Can we look at another study review to see if this is happening? My guess is that it may have to do with how the session is being stored and access in the load balanced environment. What do you think Mattison? Harry Shyket Digital Media Specialist Yale University Peabody Museum ph. 203-436-9428 har...@ya... From: William Piel [mailto:wil...@ya...] Sent: Tuesday, December 18, 2012 7:46 PM To: Mattison Ward Cc: TreeBASE devel; Shyket, Harry Subject: Reviewer access broken On Dec 7, 2012, at 10:17 AM, Mattison Ward wrote: I enabled load balancing for downloads yesterday afternoon on the production treebase.org<http://treebase.org/> site. I was able to request every study in nexml format overnight without causing any issues. Not every download completed since I used a 15 minute timeout in my download script, but the site kept responding during the entire procedure. Please let me know if you hear of any problems. Hi all, A problem seems to have cropped up: the reviewer access codes appears to be broken. These are special codes that allow data still "in progress" or "ready" to be accessed by the referees of manuscripts in review. If you follow this URL, you're able to browse through information about a submission that is "in progress": http://purl.org/phylo/treebase/phylows/study/TB2:S13742?x-access-code=dfe493384bc32ede886fdcb0ba9bb17c&format=html But if you try to download anything, you get an "Access forbidden" message. e.g.: http://purl.org/phylo/treebase/phylows/study/TB2:S13742?format=nexus&x-access-code=dfe493384bc32ede886fdcb0ba9bb17c Somehow I vaguely recall that the complexities of our access code system meant that load balancing might cause a problem. Harry: do you recall some of the issues regarding this system? I think it had to do with caching, no? regards, Bill |
|
From: William P. <wil...@ya...> - 2012-12-19 00:45:41
|
On Dec 7, 2012, at 10:17 AM, Mattison Ward wrote: > I enabled load balancing for downloads yesterday afternoon on the > production treebase.org site. I was able to request every study in > nexml format overnight without causing any issues. Not every download > completed since I used a 15 minute timeout in my download script, but > the site kept responding during the entire procedure. Please let me > know if you hear of any problems. Hi all, A problem seems to have cropped up: the reviewer access codes appears to be broken. These are special codes that allow data still "in progress" or "ready" to be accessed by the referees of manuscripts in review. If you follow this URL, you're able to browse through information about a submission that is "in progress": http://purl.org/phylo/treebase/phylows/study/TB2:S13742?x-access-code=dfe493384bc32ede886fdcb0ba9bb17c&format=html But if you try to download anything, you get an "Access forbidden" message. e.g.: http://purl.org/phylo/treebase/phylows/study/TB2:S13742?format=nexus&x-access-code=dfe493384bc32ede886fdcb0ba9bb17c Somehow I vaguely recall that the complexities of our access code system meant that load balancing might cause a problem. Harry: do you recall some of the issues regarding this system? I think it had to do with caching, no? regards, Bill |
|
From: Hilmar L. <hl...@ne...> - 2012-12-07 15:29:07
|
Excellent! -hilmar On Dec 7, 2012, at 10:17 AM, Mattison Ward wrote: > I enabled load balancing for downloads yesterday afternoon on the > production treebase.org site. I was able to request every study in > nexml format overnight without causing any issues. Not every download > completed since I used a 15 minute timeout in my download script, but > the site kept responding during the entire procedure. Please let me > know if you hear of any problems. > > Regards, > > Mattison > > ------------------------------------------------------------------------------ > LogMeIn Rescue: Anywhere, Anytime Remote support for IT. Free Trial > Remotely access PCs and mobile devices and provide instant support > Improve your efficiency, and focus on delivering more value-add services > Discover what IT Professionals Know. Rescue delivers > http://p.sf.net/sfu/logmein_12329d2d > _______________________________________________ > Treebase-devel mailing list > Tre...@li... > https://lists.sourceforge.net/lists/listinfo/treebase-devel -- =========================================================== : Hilmar Lapp -:- Durham, NC -:- informatics.nescent.org : =========================================================== |
|
From: Mattison W. <mat...@ne...> - 2012-12-07 15:18:01
|
I enabled load balancing for downloads yesterday afternoon on the production treebase.org site. I was able to request every study in nexml format overnight without causing any issues. Not every download completed since I used a 15 minute timeout in my download script, but the site kept responding during the entire procedure. Please let me know if you hear of any problems. Regards, Mattison |
|
From: William P. <wil...@ya...> - 2012-12-04 16:06:44
|
Hi Mattison, This sounds great, thanks. While load-balancing might be a problem for the Dryad push handshaking, it should not be a problem for downloads-only (e.g. the purls). bp On Dec 4, 2012, at 9:14 AM, Mattison Ward wrote: > Hi all. > > During some recent testing by the Open Tree of Life folks, I was > reminded that the Treebase site can stop responding if it gets many > rapid requests to download files in nexml format and, to a lesser > extent, in the other formats. I realized that it would take a trivial > amount of effort to load balance the Tomcat download processing to > some idle servers we have available. All other requests except > downloads would still be served by the current web server. This > avoids having to worry about session persistance and file uploads. Of > course, this does not speed up downloads of files in any format, but > it does give us a pipeline so the site can handle several concurrent > downloads without locking up. > > I have this set up now (and caching turned off for testing) on > treebase-dev.nescent.org. Hilmar has approved using the idle systems > for load balancing. Unless someone objects, I would like to enable > this load balancing of file downloads on the production site at > treebase.org on Thursday morning. > > Regards, > > Mattison |
|
From: Mattison W. <mat...@ne...> - 2012-12-04 15:17:12
|
Yes - caching will work with the load balancing. -Mattison On Tue, Dec 4, 2012 at 10:13 AM, Rutger Vos <rut...@gm...> wrote: > I think this is an excellent idea, so you certainly have my blessing. > I assume that, once the load balancing looks to be working correctly, > you will be able to turn the caching back on so we have that advantage > as well? > > On 12/4/12, Mattison Ward <mat...@ne...> wrote: >> Hi all. >> >> During some recent testing by the Open Tree of Life folks, I was >> reminded that the Treebase site can stop responding if it gets many >> rapid requests to download files in nexml format and, to a lesser >> extent, in the other formats. I realized that it would take a trivial >> amount of effort to load balance the Tomcat download processing to >> some idle servers we have available. All other requests except >> downloads would still be served by the current web server. This >> avoids having to worry about session persistance and file uploads. Of >> course, this does not speed up downloads of files in any format, but >> it does give us a pipeline so the site can handle several concurrent >> downloads without locking up. >> >> I have this set up now (and caching turned off for testing) on >> treebase-dev.nescent.org. Hilmar has approved using the idle systems >> for load balancing. Unless someone objects, I would like to enable >> this load balancing of file downloads on the production site at >> treebase.org on Thursday morning. >> >> Regards, >> >> Mattison >> >> ------------------------------------------------------------------------------ >> LogMeIn Rescue: Anywhere, Anytime Remote support for IT. Free Trial >> Remotely access PCs and mobile devices and provide instant support >> Improve your efficiency, and focus on delivering more value-add services >> Discover what IT Professionals Know. Rescue delivers >> http://p.sf.net/sfu/logmein_12329d2d >> _______________________________________________ >> Treebase-devel mailing list >> Tre...@li... >> https://lists.sourceforge.net/lists/listinfo/treebase-devel >> > > > -- > Dr. Rutger A. Vos > Bioinformaticist > NCB Naturalis > Visiting address: Office A109, Einsteinweg 2, 2333 CC, Leiden, the > Netherlands > Mailing address: Postbus 9517, 2300 RA, Leiden, the Netherlands > http://rutgervos.blogspot.com -- Mattison Ward NESCent at Duke University 2024 W. Main Street, Suite A200 Durham, NC 27705-4667 919-668-4585 (desk) 919-668-4551 (alternate) 919-668-9198 (fax) |
|
From: Mattison W. <mat...@ne...> - 2012-12-04 14:15:21
|
Hi all. During some recent testing by the Open Tree of Life folks, I was reminded that the Treebase site can stop responding if it gets many rapid requests to download files in nexml format and, to a lesser extent, in the other formats. I realized that it would take a trivial amount of effort to load balance the Tomcat download processing to some idle servers we have available. All other requests except downloads would still be served by the current web server. This avoids having to worry about session persistance and file uploads. Of course, this does not speed up downloads of files in any format, but it does give us a pipeline so the site can handle several concurrent downloads without locking up. I have this set up now (and caching turned off for testing) on treebase-dev.nescent.org. Hilmar has approved using the idle systems for load balancing. Unless someone objects, I would like to enable this load balancing of file downloads on the production site at treebase.org on Thursday morning. Regards, Mattison |
|
From: François M. <fra...@gm...> - 2012-08-14 15:38:56
|
Hi all, I agree with what Rutger and Hilmar suggest. As far as phylobase is concerned, the tree format has currently a slot to store annotations that was created with NeXML in mind. However, at this stage, it's not used by anything other than the validation of the object, so it could be used for the NeXML annotations. It would be great to complete NeXML support within NCL, as this would provide a nice way to import NeXML objects into R given that phylobase provides methods for converting objects to most commonly used tree format in R. Unfortunately I can't really help with this part given that I don't know C++ but I would be happy to write the modifications that would be needed in phylobase. Cheers, -- François On Tue, Aug 14, 2012 at 5:21 PM, Hilmar Lapp <hl...@ne...> wrote: > Hi Carl, > > I think from an interoperability and reusability standpoint, the best strategy would be to build out the support that has been started earlier in the Phylobase package. I believe the NeXML support in there started out using the XML support in R. One problem that ran into and sort of get stalled by is that the phylo4 data structure isn't yet prepared for storing arbitrary annotations, I think. > > NeXML support was started in NCL (NEXUS Class Library) in a Google Summer of Code project a few years back, but isn't complete and isn't in the main trunk. Completing that would hold lots of potential, too, though. > > I'm copying the nexml and phylobase developers lists. > > -hilmar > > On Aug 14, 2012, at 7:41 AM, Rutger Vos wrote: > >> Hi Carl, >> >> thanks for writing in! Yes, it would be great if there was NeXML >> support in R. I'm not actually an R programmer - though I'd like to >> somehow absorb that ability over time, and would be willing to >> advice/help out - but it is evident that there is a need. >> >> Ideally it would expose NeXML data in a way that is reasonably >> compatible with the different phylogenetics packages for R (ape? ouch? >> phylobase? treebase? I don't know the landscape that well). To me it >> seems that there are the following possibilities: >> >> 1. a new, simple API just for reading and writing NeXML, which other >> package authors can import >> >> 2. NeXML I/O in NCL (C++) is completed, package authors build on that, >> perhaps by way of phylobase? >> >> 3. anarchy, different authors build their own NeXML parsers, each with >> their own bugs (instead of all bugs in one place). >> >> The ideal scenario, I think, would be 2, but it would be harder than >> 1. I vote against 3. By the way, Brian O'Meara does a lot in R, it >> would be a good idea to solicit his opinion as well, especially in >> relation to the DateLife R webservice he's nurturing. >> >> Best wishes, >> >> Rutger >> >> On Tue, Aug 14, 2012 at 2:10 AM, Carl Boettiger <cbo...@gm...> wrote: >>> Dear TreeBASE developers, >>> >>> It appears that the NeXML format isn't currently supported in the R >>> environment, but offers much richer information than the old nexus format, >>> as exemplified by the treebase files under each. As I've mentioned before >>> on this list, I've written a little R package for accessing TreeBASE, which >>> relies on nexus formats for compatibility with existing R packages. I think >>> this would be a great opportunity to introduce a NeXML format reader into R, >>> and am trying to figure out where the best place is to start. I think >>> there's considerable NeXML expertise on this list, and was wondering if I >>> might find anyone else interested in helping with this effort, or give some >>> input to get me jump-started? >>> >>> I see that NeXML has C++, Java, PERL, Python, javascript and ruby >>> implementations already, and I'm wondering if the best thing to do would be >>> to wrap something like the C++ library, or start from scratch since R has a >>> pretty solid XML library already. Phylogenies in R are represented using a >>> few different and mostly very crude data structures, and we could benefit a >>> lot as R developers from a richer object -- but there's also a large >>> community of legacy formats and software relying on those formats to >>> support. I could hack something together that worked, but I think I need a >>> bit more community input to do this right. Anyone interested? >>> >>> Carl >>> >>> >>> >>> -- >>> Carl Boettiger >>> UC Davis >>> http://www.carlboettiger.info/ >>> >>> >>> ------------------------------------------------------------------------------ >>> Live Security Virtual Conference >>> Exclusive live event will cover all the ways today's security and >>> threat landscape has changed and how IT managers can respond. Discussions >>> will include endpoint security, mobile security and the latest in malware >>> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ >>> _______________________________________________ >>> Treebase-devel mailing list >>> Tre...@li... >>> https://lists.sourceforge.net/lists/listinfo/treebase-devel >>> >> >> >> >> -- >> Dr. Rutger A. Vos >> Bioinformaticist >> NCB Naturalis >> Visiting address: Office A109, Einsteinweg 2, 2333 CC, Leiden, the Netherlands >> Mailing address: Postbus 9517, 2300 RA, Leiden, the Netherlands >> http://rutgervos.blogspot.com >> >> ------------------------------------------------------------------------------ >> Live Security Virtual Conference >> Exclusive live event will cover all the ways today's security and >> threat landscape has changed and how IT managers can respond. Discussions >> will include endpoint security, mobile security and the latest in malware >> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ >> _______________________________________________ >> Treebase-devel mailing list >> Tre...@li... >> https://lists.sourceforge.net/lists/listinfo/treebase-devel > > -- > =========================================================== > : Hilmar Lapp -:- Durham, NC -:- informatics.nescent.org : > =========================================================== > > > > _______________________________________________ > Phylobase-devl mailing list > Phy...@li... > https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/phylobase-devl |
|
From: Hilmar L. <hl...@ne...> - 2012-08-14 15:21:35
|
Hi Carl, I think from an interoperability and reusability standpoint, the best strategy would be to build out the support that has been started earlier in the Phylobase package. I believe the NeXML support in there started out using the XML support in R. One problem that ran into and sort of get stalled by is that the phylo4 data structure isn't yet prepared for storing arbitrary annotations, I think. NeXML support was started in NCL (NEXUS Class Library) in a Google Summer of Code project a few years back, but isn't complete and isn't in the main trunk. Completing that would hold lots of potential, too, though. I'm copying the nexml and phylobase developers lists. -hilmar On Aug 14, 2012, at 7:41 AM, Rutger Vos wrote: > Hi Carl, > > thanks for writing in! Yes, it would be great if there was NeXML > support in R. I'm not actually an R programmer - though I'd like to > somehow absorb that ability over time, and would be willing to > advice/help out - but it is evident that there is a need. > > Ideally it would expose NeXML data in a way that is reasonably > compatible with the different phylogenetics packages for R (ape? ouch? > phylobase? treebase? I don't know the landscape that well). To me it > seems that there are the following possibilities: > > 1. a new, simple API just for reading and writing NeXML, which other > package authors can import > > 2. NeXML I/O in NCL (C++) is completed, package authors build on that, > perhaps by way of phylobase? > > 3. anarchy, different authors build their own NeXML parsers, each with > their own bugs (instead of all bugs in one place). > > The ideal scenario, I think, would be 2, but it would be harder than > 1. I vote against 3. By the way, Brian O'Meara does a lot in R, it > would be a good idea to solicit his opinion as well, especially in > relation to the DateLife R webservice he's nurturing. > > Best wishes, > > Rutger > > On Tue, Aug 14, 2012 at 2:10 AM, Carl Boettiger <cbo...@gm...> wrote: >> Dear TreeBASE developers, >> >> It appears that the NeXML format isn't currently supported in the R >> environment, but offers much richer information than the old nexus format, >> as exemplified by the treebase files under each. As I've mentioned before >> on this list, I've written a little R package for accessing TreeBASE, which >> relies on nexus formats for compatibility with existing R packages. I think >> this would be a great opportunity to introduce a NeXML format reader into R, >> and am trying to figure out where the best place is to start. I think >> there's considerable NeXML expertise on this list, and was wondering if I >> might find anyone else interested in helping with this effort, or give some >> input to get me jump-started? >> >> I see that NeXML has C++, Java, PERL, Python, javascript and ruby >> implementations already, and I'm wondering if the best thing to do would be >> to wrap something like the C++ library, or start from scratch since R has a >> pretty solid XML library already. Phylogenies in R are represented using a >> few different and mostly very crude data structures, and we could benefit a >> lot as R developers from a richer object -- but there's also a large >> community of legacy formats and software relying on those formats to >> support. I could hack something together that worked, but I think I need a >> bit more community input to do this right. Anyone interested? >> >> Carl >> >> >> >> -- >> Carl Boettiger >> UC Davis >> http://www.carlboettiger.info/ >> >> >> ------------------------------------------------------------------------------ >> Live Security Virtual Conference >> Exclusive live event will cover all the ways today's security and >> threat landscape has changed and how IT managers can respond. Discussions >> will include endpoint security, mobile security and the latest in malware >> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ >> _______________________________________________ >> Treebase-devel mailing list >> Tre...@li... >> https://lists.sourceforge.net/lists/listinfo/treebase-devel >> > > > > -- > Dr. Rutger A. Vos > Bioinformaticist > NCB Naturalis > Visiting address: Office A109, Einsteinweg 2, 2333 CC, Leiden, the Netherlands > Mailing address: Postbus 9517, 2300 RA, Leiden, the Netherlands > http://rutgervos.blogspot.com > > ------------------------------------------------------------------------------ > Live Security Virtual Conference > Exclusive live event will cover all the ways today's security and > threat landscape has changed and how IT managers can respond. Discussions > will include endpoint security, mobile security and the latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > _______________________________________________ > Treebase-devel mailing list > Tre...@li... > https://lists.sourceforge.net/lists/listinfo/treebase-devel -- =========================================================== : Hilmar Lapp -:- Durham, NC -:- informatics.nescent.org : =========================================================== |
|
From: Rutger V. <rut...@gm...> - 2012-08-14 12:38:15
|
> 2 sounds difficult (because NCL is complex), but I agree with Rutger that it is ideal. Doing that would enable NeXML reading in GARLI and RevBayes. If someone goes that route, I'm happy to answer questions and help when I can. I'm afraid that I'm a bit overcommitted right now, so I can't pledge much coding time. I'm happy to answer questions too but I don't know C++ so it would be on the NeXML end. Well, perhaps I can say useful things about what did and did not work for APIs in other languages. > On Aug 14, 2012, at 6:41 AM, Rutger Vos wrote: > >> Hi Carl, >> >> thanks for writing in! Yes, it would be great if there was NeXML >> support in R. I'm not actually an R programmer - though I'd like to >> somehow absorb that ability over time, and would be willing to >> advice/help out - but it is evident that there is a need. >> >> Ideally it would expose NeXML data in a way that is reasonably >> compatible with the different phylogenetics packages for R (ape? ouch? >> phylobase? treebase? I don't know the landscape that well). To me it >> seems that there are the following possibilities: >> >> 1. a new, simple API just for reading and writing NeXML, which other >> package authors can import >> >> 2. NeXML I/O in NCL (C++) is completed, package authors build on that, >> perhaps by way of phylobase? >> >> 3. anarchy, different authors build their own NeXML parsers, each with >> their own bugs (instead of all bugs in one place). >> >> The ideal scenario, I think, would be 2, but it would be harder than >> 1. I vote against 3. By the way, Brian O'Meara does a lot in R, it >> would be a good idea to solicit his opinion as well, especially in >> relation to the DateLife R webservice he's nurturing. >> >> Best wishes, >> >> Rutger >> >> On Tue, Aug 14, 2012 at 2:10 AM, Carl Boettiger <cbo...@gm...> wrote: >>> Dear TreeBASE developers, >>> >>> It appears that the NeXML format isn't currently supported in the R >>> environment, but offers much richer information than the old nexus format, >>> as exemplified by the treebase files under each. As I've mentioned before >>> on this list, I've written a little R package for accessing TreeBASE, which >>> relies on nexus formats for compatibility with existing R packages. I think >>> this would be a great opportunity to introduce a NeXML format reader into R, >>> and am trying to figure out where the best place is to start. I think >>> there's considerable NeXML expertise on this list, and was wondering if I >>> might find anyone else interested in helping with this effort, or give some >>> input to get me jump-started? >>> >>> I see that NeXML has C++, Java, PERL, Python, javascript and ruby >>> implementations already, and I'm wondering if the best thing to do would be >>> to wrap something like the C++ library, or start from scratch since R has a >>> pretty solid XML library already. Phylogenies in R are represented using a >>> few different and mostly very crude data structures, and we could benefit a >>> lot as R developers from a richer object -- but there's also a large >>> community of legacy formats and software relying on those formats to >>> support. I could hack something together that worked, but I think I need a >>> bit more community input to do this right. Anyone interested? >>> >>> Carl >>> >>> >>> >>> -- >>> Carl Boettiger >>> UC Davis >>> http://www.carlboettiger.info/ >>> >>> >>> ------------------------------------------------------------------------------ >>> Live Security Virtual Conference >>> Exclusive live event will cover all the ways today's security and >>> threat landscape has changed and how IT managers can respond. Discussions >>> will include endpoint security, mobile security and the latest in malware >>> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ >>> _______________________________________________ >>> Treebase-devel mailing list >>> Tre...@li... >>> https://lists.sourceforge.net/lists/listinfo/treebase-devel >>> >> >> >> >> -- >> Dr. Rutger A. Vos >> Bioinformaticist >> NCB Naturalis >> Visiting address: Office A109, Einsteinweg 2, 2333 CC, Leiden, the Netherlands >> Mailing address: Postbus 9517, 2300 RA, Leiden, the Netherlands >> http://rutgervos.blogspot.com >> >> ------------------------------------------------------------------------------ >> Live Security Virtual Conference >> Exclusive live event will cover all the ways today's security and >> threat landscape has changed and how IT managers can respond. Discussions >> will include endpoint security, mobile security and the latest in malware >> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ >> _______________________________________________ >> Treebase-devel mailing list >> Tre...@li... >> https://lists.sourceforge.net/lists/listinfo/treebase-devel > > > ------------------------------------------------------------------------------ > Live Security Virtual Conference > Exclusive live event will cover all the ways today's security and > threat landscape has changed and how IT managers can respond. Discussions > will include endpoint security, mobile security and the latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > _______________________________________________ > Treebase-devel mailing list > Tre...@li... > https://lists.sourceforge.net/lists/listinfo/treebase-devel -- Dr. Rutger A. Vos Bioinformaticist NCB Naturalis Visiting address: Office A109, Einsteinweg 2, 2333 CC, Leiden, the Netherlands Mailing address: Postbus 9517, 2300 RA, Leiden, the Netherlands http://rutgervos.blogspot.com |
|
From: Mark H. <mth...@gm...> - 2012-08-14 12:33:47
|
Hi Carl, Brian O'Meara was also a lead on phylobase's interaction with NCL, so he'd be a good contact. 2 sounds difficult (because NCL is complex), but I agree with Rutger that it is ideal. Doing that would enable NeXML reading in GARLI and RevBayes. If someone goes that route, I'm happy to answer questions and help when I can. I'm afraid that I'm a bit overcommitted right now, so I can't pledge much coding time. all the best, mark On Aug 14, 2012, at 6:41 AM, Rutger Vos wrote: > Hi Carl, > > thanks for writing in! Yes, it would be great if there was NeXML > support in R. I'm not actually an R programmer - though I'd like to > somehow absorb that ability over time, and would be willing to > advice/help out - but it is evident that there is a need. > > Ideally it would expose NeXML data in a way that is reasonably > compatible with the different phylogenetics packages for R (ape? ouch? > phylobase? treebase? I don't know the landscape that well). To me it > seems that there are the following possibilities: > > 1. a new, simple API just for reading and writing NeXML, which other > package authors can import > > 2. NeXML I/O in NCL (C++) is completed, package authors build on that, > perhaps by way of phylobase? > > 3. anarchy, different authors build their own NeXML parsers, each with > their own bugs (instead of all bugs in one place). > > The ideal scenario, I think, would be 2, but it would be harder than > 1. I vote against 3. By the way, Brian O'Meara does a lot in R, it > would be a good idea to solicit his opinion as well, especially in > relation to the DateLife R webservice he's nurturing. > > Best wishes, > > Rutger > > On Tue, Aug 14, 2012 at 2:10 AM, Carl Boettiger <cbo...@gm...> wrote: >> Dear TreeBASE developers, >> >> It appears that the NeXML format isn't currently supported in the R >> environment, but offers much richer information than the old nexus format, >> as exemplified by the treebase files under each. As I've mentioned before >> on this list, I've written a little R package for accessing TreeBASE, which >> relies on nexus formats for compatibility with existing R packages. I think >> this would be a great opportunity to introduce a NeXML format reader into R, >> and am trying to figure out where the best place is to start. I think >> there's considerable NeXML expertise on this list, and was wondering if I >> might find anyone else interested in helping with this effort, or give some >> input to get me jump-started? >> >> I see that NeXML has C++, Java, PERL, Python, javascript and ruby >> implementations already, and I'm wondering if the best thing to do would be >> to wrap something like the C++ library, or start from scratch since R has a >> pretty solid XML library already. Phylogenies in R are represented using a >> few different and mostly very crude data structures, and we could benefit a >> lot as R developers from a richer object -- but there's also a large >> community of legacy formats and software relying on those formats to >> support. I could hack something together that worked, but I think I need a >> bit more community input to do this right. Anyone interested? >> >> Carl >> >> >> >> -- >> Carl Boettiger >> UC Davis >> http://www.carlboettiger.info/ >> >> >> ------------------------------------------------------------------------------ >> Live Security Virtual Conference >> Exclusive live event will cover all the ways today's security and >> threat landscape has changed and how IT managers can respond. Discussions >> will include endpoint security, mobile security and the latest in malware >> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ >> _______________________________________________ >> Treebase-devel mailing list >> Tre...@li... >> https://lists.sourceforge.net/lists/listinfo/treebase-devel >> > > > > -- > Dr. Rutger A. Vos > Bioinformaticist > NCB Naturalis > Visiting address: Office A109, Einsteinweg 2, 2333 CC, Leiden, the Netherlands > Mailing address: Postbus 9517, 2300 RA, Leiden, the Netherlands > http://rutgervos.blogspot.com > > ------------------------------------------------------------------------------ > Live Security Virtual Conference > Exclusive live event will cover all the ways today's security and > threat landscape has changed and how IT managers can respond. Discussions > will include endpoint security, mobile security and the latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > _______________________________________________ > Treebase-devel mailing list > Tre...@li... > https://lists.sourceforge.net/lists/listinfo/treebase-devel |
|
From: Rutger V. <rut...@gm...> - 2012-08-14 11:41:15
|
Hi Carl, thanks for writing in! Yes, it would be great if there was NeXML support in R. I'm not actually an R programmer - though I'd like to somehow absorb that ability over time, and would be willing to advice/help out - but it is evident that there is a need. Ideally it would expose NeXML data in a way that is reasonably compatible with the different phylogenetics packages for R (ape? ouch? phylobase? treebase? I don't know the landscape that well). To me it seems that there are the following possibilities: 1. a new, simple API just for reading and writing NeXML, which other package authors can import 2. NeXML I/O in NCL (C++) is completed, package authors build on that, perhaps by way of phylobase? 3. anarchy, different authors build their own NeXML parsers, each with their own bugs (instead of all bugs in one place). The ideal scenario, I think, would be 2, but it would be harder than 1. I vote against 3. By the way, Brian O'Meara does a lot in R, it would be a good idea to solicit his opinion as well, especially in relation to the DateLife R webservice he's nurturing. Best wishes, Rutger On Tue, Aug 14, 2012 at 2:10 AM, Carl Boettiger <cbo...@gm...> wrote: > Dear TreeBASE developers, > > It appears that the NeXML format isn't currently supported in the R > environment, but offers much richer information than the old nexus format, > as exemplified by the treebase files under each. As I've mentioned before > on this list, I've written a little R package for accessing TreeBASE, which > relies on nexus formats for compatibility with existing R packages. I think > this would be a great opportunity to introduce a NeXML format reader into R, > and am trying to figure out where the best place is to start. I think > there's considerable NeXML expertise on this list, and was wondering if I > might find anyone else interested in helping with this effort, or give some > input to get me jump-started? > > I see that NeXML has C++, Java, PERL, Python, javascript and ruby > implementations already, and I'm wondering if the best thing to do would be > to wrap something like the C++ library, or start from scratch since R has a > pretty solid XML library already. Phylogenies in R are represented using a > few different and mostly very crude data structures, and we could benefit a > lot as R developers from a richer object -- but there's also a large > community of legacy formats and software relying on those formats to > support. I could hack something together that worked, but I think I need a > bit more community input to do this right. Anyone interested? > > Carl > > > > -- > Carl Boettiger > UC Davis > http://www.carlboettiger.info/ > > > ------------------------------------------------------------------------------ > Live Security Virtual Conference > Exclusive live event will cover all the ways today's security and > threat landscape has changed and how IT managers can respond. Discussions > will include endpoint security, mobile security and the latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > _______________________________________________ > Treebase-devel mailing list > Tre...@li... > https://lists.sourceforge.net/lists/listinfo/treebase-devel > -- Dr. Rutger A. Vos Bioinformaticist NCB Naturalis Visiting address: Office A109, Einsteinweg 2, 2333 CC, Leiden, the Netherlands Mailing address: Postbus 9517, 2300 RA, Leiden, the Netherlands http://rutgervos.blogspot.com |
|
From: Carl B. <cbo...@gm...> - 2012-08-13 22:10:28
|
Dear TreeBASE developers, It appears that the NeXML format isn't currently supported in the R environment, but offers much richer information than the old nexus format, as exemplified by the treebase files under each. As I've mentioned before on this list, I've written a little R package for accessing TreeBASE, which relies on nexus formats for compatibility with existing R packages. I think this would be a great opportunity to introduce a NeXML format reader into R, and am trying to figure out where the best place is to start. I think there's considerable NeXML expertise on this list, and was wondering if I might find anyone else interested in helping with this effort, or give some input to get me jump-started? I see that NeXML <http://nexml.org> has C++, Java, PERL, Python, javascript and ruby implementations already, and I'm wondering if the best thing to do would be to wrap something like the C++ library, or start from scratch since R has a pretty solid XML library already. Phylogenies in R are represented using a few different and mostly very crude data structures, and we could benefit a lot as R developers from a richer object -- but there's also a large community of legacy formats and software relying on those formats to support. I could hack something together that worked, but I think I need a bit more community input to do this right. Anyone interested? Carl -- Carl Boettiger UC Davis http://www.carlboettiger.info/ |
|
From: William P. <wil...@ya...> - 2012-08-07 15:18:39
|
Sorry everyone -- it looks like this was a case of my hotel's router doing some faulty excessive caching… bp On Aug 7, 2012, at 9:28 PM, Shyket, Harry wrote: > Could this be an issue with the move to load balancing? It could be that it is creating new session ids moving from logged out to logged in and vice versa. That may send it off to a different TreeBASE instance. > > If so, then we should consider moving the session data to the database. > > Harry Shyket |
|
From: Mattison W. <mat...@ne...> - 2012-08-07 13:38:49
|
I haven't started to set up load balancing - we are still on a single tomcat server for now. On Tue, Aug 7, 2012 at 9:28 AM, Shyket, Harry <har...@ya...> wrote: > Could this be an issue with the move to load balancing? It could be that > it is creating new session ids moving from logged out to logged in and vice > versa. That may send it off to a different TreeBASE instance. > > > > If so, then we should consider moving the session data to the database. > > > > Harry Shyket > > Digital Media Specialist > > Yale University Peabody Museum > > ph. 203-436-9428 > > har...@ya... > > > > From: William Piel [mailto:wil...@ya...] > Sent: Sunday, August 05, 2012 9:17 AM > To: NESCent Issue Tracker (help) > Cc: TreeBASE devel; Michael Jones > Subject: [Treebase-devel] TreeBASE malfunction > > > > > > Something bizarre seems to be happening to TreeBASE -- it was not a problem > as late as 7/30, but it seems to be a problem as of 8/5. I'm not sure what > happened between 7/30 and 8/5, but I'm now getting the following new > symptoms: > > > > - Normally, when I click the "log out" link, I can log out. Now it still > behaves as though I'm logged in even after clicking "log out". > > > > - With the "Submit" button, I am normally required to log in, and then after > successfully logging in I'm normally directed to a list of my existing > submissions. Now after logging in I'm immediately presented with the login > screen again. (but if I click the "Search" button I'm offered a > "Submissions" tab, so I can get to this that way). > > > > - Normally, if I'm logged in and I click the "Citations" link, I get to see > the citation metadata for my submission, now I get another "log in" screen > regardless of the fact that I'm already logged in. > > > > Emptying my local cache and blocking browser caching does not solve the > problem. Switching from Safari to FireFox, Camino, etc, also solves nothing. > > > > My sense is that this problem comes from server-side caching that's gone out > of control. The only caching we should do is for links that pull data, e.g.: > > > > http://purl.org/phylo/treebase/phylows/study/TB2:S1000?format=nexml > > http://purl.org/phylo/treebase/phylows/study/TB2:S1000?format=nexus > > http://purl.org/phylo/treebase/phylows/matrix/TB2:M553?format=nexus > > http://purl.org/phylo/treebase/phylows/matrix/TB2:M553?format=nexml > > http://purl.org/phylo/treebase/phylows/tree/TB2:Tr4105?format=nexus > > http://purl.org/phylo/treebase/phylows/tree/TB2:Tr4105?format=nexml > > > > … which I think means redirecting to URLs that look like: > > > > /treebase-web/search/downloadAStudy.html > > /treebase-web/search/study/anyObjectAsRDF.rdf > > /treebase-web/search/study/summary.html > > /treebase-web/search/downloadANexusFile.html > > /treebase-web/search/downloadAMatrix.html > > /treebase-web/search/downloadATree.html > > /treebase-web/search/downloadAnAnalysisStep.html > > > > But we don't want it to cache these: > > > > http://treebase.org/treebase-web/search/studySearch.html > > http://treebase.org/treebase-web/search/treeSearch.html > > http://treebase.org/treebase-web/search/matrixSearch.html > > http://treebase.org/treebase-web/search/taxonSearch.html > > > > .. because the results are unstable. Also, we don't want to cache the user > and admin pages, such as these: > > > > http://treebase.org/treebase-web/user/submissionList.html > > http://treebase.org/treebase-web/user/citationForm.html > > (etc) > > > > or these: > > > > http://treebase.org/treebase-web/admin/administrationPage.html > > http://treebase.org/treebase-web/admin/readyStateStudies.html > > (etc) > > > > bp > > > > > > > > > > > > > ------------------------------------------------------------------------------ > Live Security Virtual Conference > Exclusive live event will cover all the ways today's security and > threat landscape has changed and how IT managers can respond. Discussions > will include endpoint security, mobile security and the latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > _______________________________________________ > Treebase-devel mailing list > Tre...@li... > https://lists.sourceforge.net/lists/listinfo/treebase-devel > -- Mattison Ward NESCent at Duke University 2024 W. Main Street, Suite A200 Durham, NC 27705-4667 919-668-4585 (desk) 919-668-4551 (alternate) 919-668-9198 (fax) |
|
From: Shyket, H. <har...@ya...> - 2012-08-07 13:28:40
|
Could this be an issue with the move to load balancing? It could be that it is creating new session ids moving from logged out to logged in and vice versa. That may send it off to a different TreeBASE instance. If so, then we should consider moving the session data to the database. Harry Shyket Digital Media Specialist Yale University Peabody Museum ph. 203-436-9428 har...@ya... From: William Piel [mailto:wil...@ya...] Sent: Sunday, August 05, 2012 9:17 AM To: NESCent Issue Tracker (help) Cc: TreeBASE devel; Michael Jones Subject: [Treebase-devel] TreeBASE malfunction Something bizarre seems to be happening to TreeBASE -- it was not a problem as late as 7/30, but it seems to be a problem as of 8/5. I'm not sure what happened between 7/30 and 8/5, but I'm now getting the following new symptoms: - Normally, when I click the "log out" link, I can log out. Now it still behaves as though I'm logged in even after clicking "log out". - With the "Submit" button, I am normally required to log in, and then after successfully logging in I'm normally directed to a list of my existing submissions. Now after logging in I'm immediately presented with the login screen again. (but if I click the "Search" button I'm offered a "Submissions" tab, so I can get to this that way). - Normally, if I'm logged in and I click the "Citations" link, I get to see the citation metadata for my submission, now I get another "log in" screen regardless of the fact that I'm already logged in. Emptying my local cache and blocking browser caching does not solve the problem. Switching from Safari to FireFox, Camino, etc, also solves nothing. My sense is that this problem comes from server-side caching that's gone out of control. The only caching we should do is for links that pull data, e.g.: http://purl.org/phylo/treebase/phylows/study/TB2:S1000?format=nexml http://purl.org/phylo/treebase/phylows/study/TB2:S1000?format=nexus http://purl.org/phylo/treebase/phylows/matrix/TB2:M553?format=nexus http://purl.org/phylo/treebase/phylows/matrix/TB2:M553?format=nexml http://purl.org/phylo/treebase/phylows/tree/TB2:Tr4105?format=nexus http://purl.org/phylo/treebase/phylows/tree/TB2:Tr4105?format=nexml ... which I think means redirecting to URLs that look like: /treebase-web/search/downloadAStudy.html /treebase-web/search/study/anyObjectAsRDF.rdf /treebase-web/search/study/summary.html /treebase-web/search/downloadANexusFile.html /treebase-web/search/downloadAMatrix.html /treebase-web/search/downloadATree.html /treebase-web/search/downloadAnAnalysisStep.html But we don't want it to cache these: http://treebase.org/treebase-web/search/studySearch.html http://treebase.org/treebase-web/search/treeSearch.html http://treebase.org/treebase-web/search/matrixSearch.html http://treebase.org/treebase-web/search/taxonSearch.html .. because the results are unstable. Also, we don't want to cache the user and admin pages, such as these: http://treebase.org/treebase-web/user/submissionList.html http://treebase.org/treebase-web/user/citationForm.html (etc) or these: http://treebase.org/treebase-web/admin/administrationPage.html http://treebase.org/treebase-web/admin/readyStateStudies.html (etc) bp |
|
From: William P. <wil...@ya...> - 2012-08-05 13:17:37
|
Something bizarre seems to be happening to TreeBASE -- it was not a problem as late as 7/30, but it seems to be a problem as of 8/5. I'm not sure what happened between 7/30 and 8/5, but I'm now getting the following new symptoms: - Normally, when I click the "log out" link, I can log out. Now it still behaves as though I'm logged in even after clicking "log out". - With the "Submit" button, I am normally required to log in, and then after successfully logging in I'm normally directed to a list of my existing submissions. Now after logging in I'm immediately presented with the login screen again. (but if I click the "Search" button I'm offered a "Submissions" tab, so I can get to this that way). - Normally, if I'm logged in and I click the "Citations" link, I get to see the citation metadata for my submission, now I get another "log in" screen regardless of the fact that I'm already logged in. Emptying my local cache and blocking browser caching does not solve the problem. Switching from Safari to FireFox, Camino, etc, also solves nothing. My sense is that this problem comes from server-side caching that's gone out of control. The only caching we should do is for links that pull data, e.g.: http://purl.org/phylo/treebase/phylows/study/TB2:S1000?format=nexml http://purl.org/phylo/treebase/phylows/study/TB2:S1000?format=nexus http://purl.org/phylo/treebase/phylows/matrix/TB2:M553?format=nexus http://purl.org/phylo/treebase/phylows/matrix/TB2:M553?format=nexml http://purl.org/phylo/treebase/phylows/tree/TB2:Tr4105?format=nexus http://purl.org/phylo/treebase/phylows/tree/TB2:Tr4105?format=nexml … which I think means redirecting to URLs that look like: /treebase-web/search/downloadAStudy.html /treebase-web/search/study/anyObjectAsRDF.rdf /treebase-web/search/study/summary.html /treebase-web/search/downloadANexusFile.html /treebase-web/search/downloadAMatrix.html /treebase-web/search/downloadATree.html /treebase-web/search/downloadAnAnalysisStep.html But we don't want it to cache these: http://treebase.org/treebase-web/search/studySearch.html http://treebase.org/treebase-web/search/treeSearch.html http://treebase.org/treebase-web/search/matrixSearch.html http://treebase.org/treebase-web/search/taxonSearch.html .. because the results are unstable. Also, we don't want to cache the user and admin pages, such as these: http://treebase.org/treebase-web/user/submissionList.html http://treebase.org/treebase-web/user/citationForm.html (etc) or these: http://treebase.org/treebase-web/admin/administrationPage.html http://treebase.org/treebase-web/admin/readyStateStudies.html (etc) bp |
|
From: Mattison W. <mat...@ne...> - 2012-07-25 15:00:37
|
If the file isn't inserted into the database, it would only be accessible from the instance it is copied to. When the "delivered file unpacked and parked" is it inserted into the database? On Wed, Jul 25, 2012 at 10:50 AM, Shyket, Harry <har...@ya...> wrote: > The Dryad issue comes down to whether the files are all synced across the > multiple instances of Tomcat. > > > > Mattison, > > > > What happens when a file is created on one TreeBASE instance, but then there > is an attempt to access it from another? > > > > Thanks, > > > > Harry Shyket > > Digital Media Specialist > > Yale University Peabody Museum > > ph. 203-436-9428 > > har...@ya... > > > > From: William Piel [mailto:wil...@ya...] > Sent: Wednesday, July 25, 2012 10:44 AM > To: Mattison Ward > Cc: TreeBASE devel > Subject: Re: [Treebase-devel] load balancing treebase? > > > > > > On Jul 25, 2012, at 9:36 AM, Mattison Ward wrote: > > > > The web server would load balance by session such that a session would > always be handled by the same tomcat instance (unless the instance > failed). > > > > Much of it should work okay -- e.g. seeing as uploaded files are stored as > blobs in the database, so should be accessible by all tomcat instances. > > > > Harry can correct me if I'm wrong -- but I think that one place where this > might not work is in the handshaking between Dryad and TreeBASE: files get > pushed at one time point while the user is logged into Dryad, an email with > a link to TreeBASE is issued, then at a later point the user follows the > email link and the delivered file is unpacked and parked into the user's > TreeBASE account. So there might be a problem with files pushed from Dryad > winding up in the directory of a different instance from what the user later > uses to try to retrieve it. > > > > bp > > > > -- Mattison Ward NESCent at Duke University 2024 W. Main Street, Suite A200 Durham, NC 27705-4667 919-668-4585 (desk) 919-668-4551 (alternate) 919-668-9198 (fax) |
|
From: Shyket, H. <har...@ya...> - 2012-07-25 14:51:11
|
The Dryad issue comes down to whether the files are all synced across the multiple instances of Tomcat. Mattison, What happens when a file is created on one TreeBASE instance, but then there is an attempt to access it from another? Thanks, Harry Shyket Digital Media Specialist Yale University Peabody Museum ph. 203-436-9428 har...@ya... From: William Piel [mailto:wil...@ya...] Sent: Wednesday, July 25, 2012 10:44 AM To: Mattison Ward Cc: TreeBASE devel Subject: Re: [Treebase-devel] load balancing treebase? On Jul 25, 2012, at 9:36 AM, Mattison Ward wrote: The web server would load balance by session such that a session would always be handled by the same tomcat instance (unless the instance failed). Much of it should work okay -- e.g. seeing as uploaded files are stored as blobs in the database, so should be accessible by all tomcat instances. Harry can correct me if I'm wrong -- but I think that one place where this might not work is in the handshaking between Dryad and TreeBASE: files get pushed at one time point while the user is logged into Dryad, an email with a link to TreeBASE is issued, then at a later point the user follows the email link and the delivered file is unpacked and parked into the user's TreeBASE account. So there might be a problem with files pushed from Dryad winding up in the directory of a different instance from what the user later uses to try to retrieve it. bp |
|
From: William P. <wil...@ya...> - 2012-07-25 14:44:14
|
On Jul 25, 2012, at 9:36 AM, Mattison Ward wrote: > The web server would load balance by session such that a session would > always be handled by the same tomcat instance (unless the instance > failed). Much of it should work okay -- e.g. seeing as uploaded files are stored as blobs in the database, so should be accessible by all tomcat instances. Harry can correct me if I'm wrong -- but I think that one place where this might not work is in the handshaking between Dryad and TreeBASE: files get pushed at one time point while the user is logged into Dryad, an email with a link to TreeBASE is issued, then at a later point the user follows the email link and the delivered file is unpacked and parked into the user's TreeBASE account. So there might be a problem with files pushed from Dryad winding up in the directory of a different instance from what the user later uses to try to retrieve it. bp |
9 messages has been excluded from this view by a project administrator.
