Windows S5R2 App 4.28 available for Beta Test

Bernd Machenschalk
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 4265
Credit: 244922893
RAC: 16808
Topic 192967

A new Windows App is available from our Beta Test Page.

This App is part of a new generation of Apps that follow our attempts to track down and fix the remaining problems of the new software we're using since S5R2 (There will soon be a MacOS PPC App once I've fixed a CPU-specific problem there).

This may remain an intermediate release that will never make it to become official Apps if we get enough feedback from running the Beta Test Apps that we can fix some more problems for the next "official" Apps.

Following a suggestion from Gary these Beta test packages feature an app_info.xml that should allow for continuing a Task that has been assigned to and possibly partially crunched with an older App version with this Beta App (keeping the old App version number, though). Note that there are some clients with which this doesn't work as desired, i.e. that reset the project when switching the platform (I've seen this behavior with the first Universal (PPC+Intel) MacOS Clients). So this Beta Test also covers this app_info.xml scheme. If in doubt, wait at least until the currently running Task has finished, report it manually and install the Beta App then.

gcc-compiled Apps such as the ones from MacOS and Linux apparently get a little speedup from using ftol() instead of modf() (which we originally did to fix problems of the Windows App), but this release is not primarily meant to improve speed. We'll work on this not before we have the validation problems found solved.

Please test&report here.

BM

BM

Gary Roberts
Gary Roberts
Moderator
Joined: 9 Feb 05
Posts: 5842
Credit: 109385356127
RAC: 35932209

Windows S5R2 App 4.28 available for Beta Test

Quote:

Following a suggestion from Gary these Beta test packages feature an app_info.xml that should allow for continuing a Task that has been assigned to and possibly partially crunched with an older App version ....

I can confirm that the app_info.xml that is supplied with this beta test does work as advertised. I have immediately switched two AMD/Windows machines to the 4.28 beta test app without incident. Both machines have 3 day caches which will now be crunched with the new app even though the tasks still show as 4.24 under the tasks tab of Boinc Manager. Windows task manager shows that 4.28 is running.

Both machines had partially completed jobs (92% and 94% done) which will be finished with the new app. There has been no problem with the restart under the new app. This was also my experience when I did this for the transition from 4.17 to 4.24

For any Windows users who would like to participate in the beta test immediately, this is what you need to do. You can make your own decisions about things like backing up, etc ... You also assume all risks if something goes wrong.

This is exactly what I did and in the exact order:-

  • *Extract the app_info.xml and the 4.28 executable from the archive
    *Copy and paste both into the BOINC\\projects\\einstein.phys.uwm.edu\\ folder
    *Do not touch any other file in that folder
    *Completely stop boinc
    *Restart boinc

The messages tab of Boinc Manager will say that the result in progress is being restarted with 424 but in actual fact 4.28 really is being used.

In about two hours time both my test results will finish and I can already see that one of them has a wingman that has already finished. Hopefully that one will validate straight away.

Cheers,
Gary.

Svenie25
Svenie25
Joined: 21 Mar 05
Posts: 139
Credit: 2436862
RAC: 0

Okay, seems to be easy to

Okay, seems to be easy to change to Betatest. ;) I will try it, too.

Gary Roberts
Gary Roberts
Moderator
Joined: 9 Feb 05
Posts: 5842
Credit: 109385356127
RAC: 35932209

There is one interesting

There is one interesting consequence of doing what I have just posted. It happened on both machines so I'm assuming it is a consequence of the new style app_info.xml. What I'm now describing didn't happen when I did this for the 4.17 to 4.24 transition.

To illustrate what happened, here is an extract from the startup messages:-

Quote:

16/07/2007 8:28:56 PM||Starting BOINC client version 5.8.15 for windows_intelx86
16/07/2007 8:28:56 PM||log flags: task, file_xfer, sched_ops
16/07/2007 8:28:56 PM||Libraries: libcurl/7.16.0 OpenSSL/0.9.8a zlib/1.2.3
16/07/2007 8:28:56 PM||Executing as a daemon
16/07/2007 8:28:56 PM||Data directory: D:\\Program Files\\BOINC
16/07/2007 8:28:56 PM|Einstein@Home|Found app_info.xml; using anonymous platform
16/07/2007 8:28:56 PM||Processor: 1 AuthenticAMD AMD Athlon(tm) XP [x86 Family 6 Model 8 Stepping 1] [fpu tsc sse 3dnow mmx]
16/07/2007 8:28:56 PM||Memory: 511.48 MB physical, 1.10 GB virtual
16/07/2007 8:28:56 PM||Disk: 19.53 GB total, 7.47 GB free
16/07/2007 8:28:57 PM|Einstein@Home|URL: http://einstein.phys.uwm.edu/; Computer ID: 12594; location: school; project prefs: school
16/07/2007 8:28:57 PM||General prefs: from Einstein@Home (last modified 2007-07-01 13:10:53)
16/07/2007 8:28:57 PM||Host location: school
16/07/2007 8:28:57 PM||General prefs: using separate prefs for school
16/07/2007 8:28:57 PM||file projects/einstein.phys.uwm.edu/einstein_S5R2_4.17_windows_intelx86.exe not found
16/07/2007 8:28:57 PM||file projects/einstein.phys.uwm.edu/einstein_S5R2_4.17_windows_intelx86.pdb not found
16/07/2007 8:28:57 PM|Einstein@Home|Restarting task h1_0463.35_S5R2__100_S5R2c_3 using einstein_S5R2 version 424

The bit I'm talking about is the info message that 4.17 versions couldn't be found.

In fact, when I look in the project folder *all* previous executables seem to have disappeared. Even the 4.24 app is gone so it looks like there's no going back :).

It happened on both machines and after I saw it on the first one I confirmed the contents before starting the new app on the second box. I had really old versions going back to ones like Albert 4.37 for example - in fact every app ever released since the project opened its doors back in Feb 2005 :).

Suddenly I have a bunch of freed up disk space so I'm not complaining :).

I've still got all the l1_* data files (so that hasn't been sorted yet) but at least the old executables are gone.

Hey Bernd, maybe you'd like to clue us in on how you performed this bit of magic. I can't really see anything obvious in the app_info.xml so I guess it has to be some other trick you've pulled :).

Cheers,
Gary.

josep
josep
Joined: 9 Mar 05
Posts: 63
Credit: 1156542
RAC: 0

The same seems to happen with


The same seems to happen with the linux version. The previous 4.21 app has been automatically deleted from my project folder.

Now only the beta app is there.

Bernd Machenschalk
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 4265
Credit: 244922893
RAC: 16808

RE: There is one

Message 70087 in response to message 70085

Quote:

There is one interesting consequence of doing what I have just posted.

Quote:

16/07/2007 8:28:57 PM||file projects/einstein.phys.uwm.edu/einstein_S5R2_4.17_windows_intelx86.exe not found
16/07/2007 8:28:57 PM||file projects/einstein.phys.uwm.edu/einstein_S5R2_4.17_windows_intelx86.pdb not found

In fact, when I look in the project folder *all* previous executables seem to have disappeared. Even the 4.24 app is gone so it looks like there's no going back :).

I've still got all the l1_* data files (so that hasn't been sorted yet) but at least the old executables are gone.


Well, the BOINC client is rather eager to delete files that he thinks are no longer needed. I thought that the file entries at the beginning of the app_info.xml should prevent the older App versions from being deleted (and just issue the warning you mentioned if they have already been deleted earlier), but apparently that's not the case; in fact it looks like it has the opposite effect.

As for the l1_* files:

Unfortunately we can't issue delete requests automatically (by the scheduler(!)) for l1_ files where a single workunit is still in the database that uses that file, which makes this cleanup rather slow (checking whether it's this particular host that got a WU that blocks this file is too slow for a scheduler request).

If the disk gets filled to the point where the host gets no work because of diskspace, the scheduler should send a request to delete the present datafiles, so the old l1_* files will be deleted this way, too.

To manually free up diskspace it should be safe to delete those l1_ files that don't have a matching h1_ file in the same directory.

Though actually the cleanup is a feature of the BOINC Client triggered by the scheduler, it sounds like an interesting idea to put some cleanup feature into the App for this case. I'll think about it.

BM

BM

Gary Roberts
Gary Roberts
Moderator
Joined: 9 Feb 05
Posts: 5842
Credit: 109385356127
RAC: 35932209

My two windows boxes that

My two windows boxes that were switched to the beta test have now successfully finished the results with the new app. Here is the quorum of the one that had a "finished" wingman and as you can see, the result validated correctly. The other is pending but looks OK so far.

I've now got 2 Linux boxes on the new app as well and all looks fine. I think I'll wait and see for a bit before switching more :).

Cheers,
Gary.

Stick
Stick
Joined: 24 Feb 05
Posts: 790
Credit: 31192054
RAC: 415

RE: My two windows boxes

Message 70089 in response to message 70088

Quote:
My two windows boxes that were switched to the beta test have now successfully finished the results with the new app. Here is the quorum of the one that had a "finished" wingman and as you can see, the result validated correctly. The other is pending but looks OK so far.

Gary,

Looking at the results of wuid=34104604 which you referred to above, I see one v4.17 result and one v4.24 result, but none from v4.28. Is this a reporting error or were you premature in your posting?

Stick

Bernd Machenschalk
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 4265
Credit: 244922893
RAC: 16808

RE: Looking at the results

Message 70090 in response to message 70089

Quote:
Looking at the results of wuid=34104604 which you referred to above, I see one v4.17 result and one v4.24 result, but none from v4.28. Is this a reporting error or were you premature in your posting?


That's a (desired) effect of the app_info.xml. Tasks that are interrupted and crunched halfway with an older App will keep the old number, only Tasks that have been ran completely with the new App will show up with the new version number.

Note that the new App doesn't do anything with the calculations that have already been done; it might even be that "mixed" results are less likely to validate correctly than "pure" ones from the old App, though these cases should be rather rare. Another edit: thinking about the changes in the App this shouldn't apply here at all - there should be no such case with this new App.

BM

BM

Gary Roberts
Gary Roberts
Moderator
Joined: 9 Feb 05
Posts: 5842
Credit: 109385356127
RAC: 35932209

Take a look more closely (ie

Take a look more closely (ie scroll through it) and you will see where the changeover to 4.28 occurred. Because the result was "branded" in the state file as 4.24, I think it still reports that way. That's OK as well over 90% was done by 4.24 anyway :).

Cheers,
Gary.

Svenie25
Svenie25
Joined: 21 Mar 05
Posts: 139
Credit: 2436862
RAC: 0

Okay, first WU particular

Okay, first WU particular finished with 4.28 is ready and validated against one of Bruce Allens Linuxhosts. Seemd to be no problems up to now.

http://einsteinathome.org/workunit/34248843

I also looked the graphics, worked fine, too. Lets wait if there are no problems in the future.

MfG
Sven S. aus G.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.