Ghost WU issue (and some talk about deadlines)

Message boards : Number crunching : Ghost WU issue (and some talk about deadlines)
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 9 · 10 · 11 · 12

AuthorMessage
Profile KWSN - Chicken of Angnor
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 9 Jul 99
Posts: 1199
Credit: 6,615,780
RAC: 0
Austria
Message 572382 - Posted: 20 May 2007, 18:55:09 UTC - in response to Message 572343.  

Simon,

I think he means platforms like OS/2 Warp, HP-UX etc. - the ones that anonymous platform was designed to run, for self-compile or user-group compile.

And Linux 2.6.16-20061110.ydl.2ps3, which is what he's using.

No, I'm afraid I haven't seen anyone post a work-round for them.

I have a feeling you're probably right about that, on second thought. I "solved" my 64-bit BOINC problem by reverting to 32-bit with a 64-bit science app (which does work with the workarounds)...

However, that's not a solution for most people running nonstandard platforms, x86_64 has it easy compared.

Regards,
Simon.
Donate to SETI@Home via PayPal!

Optimized SETI@Home apps + Information
ID: 572382 · Report as offensive
Odysseus
Volunteer tester
Avatar

Send message
Joined: 26 Jul 99
Posts: 1808
Credit: 6,701,347
RAC: 6
Canada
Message 572445 - Posted: 20 May 2007, 20:26:17 UTC - in response to Message 572211.  

Yes, but just the two entries: one for v5.13 and one for v5.17 because of the ‘crossed wires’ with Beta.

That bug was fixed with 5.8.x, I believe. I haven't had that entry since upgrading to 5.8.x a while back, and everything works fine.

I’m running v5.4.9 on this host.
ID: 572445 · Report as offensive
mdpagel

Send message
Joined: 18 Sep 99
Posts: 53
Credit: 2,619,543
RAC: 0
United States
Message 572949 - Posted: 21 May 2007, 17:33:33 UTC

I have tried the following thusfar:
redownload boinc and apply over the top of existing installation <--- still didn't download new workunits
Replace KWSN app with new version <--- still no download
Delete BOINC directory <-- failed to delete (windows complained)
Add/Remove programs on BOINC <-- didn't completely get rid of BOINC programs
Create new directory in BOINC directory and move all files into that directory <-- seemed to work
Re-install BOINC and re-attach <-- worked, did benchmarks, but still not downloading WUs. Futher, there is no client in my s@h directory (C:\\Program Files\\BOINC\\projects\\setiathome.berkeley.edu) - the directory is completely empty
Re-install optimized client <--- it's in the directory now at least;)

Currently statistics shows that I've completed no work ever, etc.

Boinc is a service installation on this computer.

I had gotten about 20 new WUs after the thumper issues, but after a while it stopped downloading additional units. I've gotten HTTP internal server errors, as well as Scheduler request failed: server returned nothing (no headers, no data), Access to reference site succeeded - project servers may be temporarily down. My "results" show that I have several workunits that seti thinks I've downloaded but haven't processed. I've never seen these WUs (before or after uninstall).

ID: 572949 · Report as offensive
Profile Geek@Play
Volunteer tester
Avatar

Send message
Joined: 31 Jul 01
Posts: 2467
Credit: 86,146,931
RAC: 0
United States
Message 573009 - Posted: 21 May 2007, 18:15:19 UTC - in response to Message 572949.  

I have tried the following thusfar:
redownload boinc and apply over the top of existing installation <--- still didn't download new workunits
Replace KWSN app with new version <--- still no download
Delete BOINC directory <-- failed to delete (windows complained)
Add/Remove programs on BOINC <-- didn't completely get rid of BOINC programs
Create new directory in BOINC directory and move all files into that directory <-- seemed to work
Re-install BOINC and re-attach <-- worked, did benchmarks, but still not downloading WUs. Futher, there is no client in my s@h directory (C:\\Program Files\\BOINC\\projects\\setiathome.berkeley.edu) - the directory is completely empty
Re-install optimized client <--- it's in the directory now at least;)

Currently statistics shows that I've completed no work ever, etc.

Boinc is a service installation on this computer.

I had gotten about 20 new WUs after the thumper issues, but after a while it stopped downloading additional units. I've gotten HTTP internal server errors, as well as Scheduler request failed: server returned nothing (no headers, no data), Access to reference site succeeded - project servers may be temporarily down. My "results" show that I have several workunits that seti thinks I've downloaded but haven't processed. I've never seen these WUs (before or after uninstall).



You must have created a new account rather than using your old account.


Boinc....Boinc....Boinc....Boinc....
ID: 573009 · Report as offensive
Profile Philadelphia
Volunteer tester
Avatar

Send message
Joined: 12 Feb 07
Posts: 1590
Credit: 399,688
RAC: 0
United States
Message 573172 - Posted: 21 May 2007, 20:59:50 UTC

Any idea how long it takes for the ghosts to disappear after doing the detach/reattach? Just curious - mine are still there as "client detached" I was able to do it on my second computer this morning so both should be clean of ghosts after the 'client detached'(ghosts) go away.
ID: 573172 · Report as offensive
Profile Jason Safoutin
Volunteer tester
Avatar

Send message
Joined: 8 Sep 05
Posts: 1386
Credit: 200,389
RAC: 0
United States
Message 573177 - Posted: 21 May 2007, 21:05:07 UTC - in response to Message 573172.  
Last modified: 21 May 2007, 21:06:04 UTC

Any idea how long it takes for the ghosts to disappear after doing the detach/reattach? Just curious - mine are still there as "client detached" I was able to do it on my second computer this morning so both should be clean of ghosts after the 'client detached'(ghosts) go away.


Just looked at mine...though I got rid of them...nope...But good news is I only have about 7 or 8 ghosts...I am just going to let the WU's run their time out rather re attach...too much trouble IMO.

EDIT: Nevermind, All WU's accounted for...Had more work than I thought...no ghosts here and I never re attached.
"By faith we understand that the universe was formed at God's command, so that what is seen was not made out of what was visible". Hebrews 11.3

ID: 573177 · Report as offensive
Profile Richard Williams

Send message
Joined: 14 Jan 04
Posts: 10
Credit: 101,524
RAC: 0
United States
Message 573186 - Posted: 21 May 2007, 21:15:00 UTC

I have to be honest, I don't know that I am buying any of the conjectured explanations. I have four boxes that for two weeks now have been not getting WU's sent to them because they keep being told there isn't any.

Between 5/11 and 5/17, my boxes got zero WU's. Then, on 5/18 I shut down BOINC on them all, and then rebooted all of them. Suddenly I got 356 credits on 5/18, still only 1/3 of my normal workload.

But then I got jst 60 on 5/19, and zero again on 5/20.

Again, I shut down BOINC and then rebooted all the machines. Voila, it now shows 437 credits on 5/21. But I look at this laptop, for instance, and BOINC is showing an error, that there is no work .... that is, until I shut down BOINC again of course, then I see this.

5/21/2007 1:45:20 PM|SETI@home|Fetching scheduler list
5/21/2007 1:45:25 PM|SETI@home|Master file download succeeded
5/21/2007 1:45:31 PM|SETI@home|Sending scheduler request: To fetch work
5/21/2007 1:45:31 PM|SETI@home|Requesting 8640 seconds of new work
5/21/2007 1:45:58 PM|SETI@home|Scheduler RPC succeeded [server version 509]
5/21/2007 1:45:58 PM|SETI@home|Deferring communication for 11 sec
5/21/2007 1:45:58 PM|SETI@home|Reason: requested by project
5/21/2007 1:45:58 PM|SETI@home|Deferring communication for 1 min 0 sec
5/21/2007 1:45:58 PM|SETI@home|Reason: no work from project
5/21/2007 1:47:02 PM|SETI@home|Sending scheduler request: To fetch work
5/21/2007 1:47:02 PM|SETI@home|Requesting 8640 seconds of new work
5/21/2007 1:47:29 PM|SETI@home|Scheduler RPC succeeded [server version 509]
5/21/2007 1:47:29 PM|SETI@home|Deferring communication for 11 sec
5/21/2007 1:47:29 PM|SETI@home|Reason: requested by project
5/21/2007 1:47:29 PM|SETI@home|Deferring communication for 1 min 0 sec
5/21/2007 1:47:29 PM|SETI@home|Reason: no work from project
5/21/2007 1:48:34 PM|SETI@home|Sending scheduler request: To fetch work
5/21/2007 1:48:34 PM|SETI@home|Requesting 8640 seconds of new work
5/21/2007 1:49:22 PM|SETI@home|Scheduler RPC succeeded [server version 509]
5/21/2007 1:49:22 PM|SETI@home|Deferring communication for 11 sec
5/21/2007 1:49:22 PM|SETI@home|Reason: requested by project
5/21/2007 1:49:25 PM|SETI@home|[file_xfer] Started download of file 05mr05aa.1632.21490.4826.3.200
5/21/2007 1:50:18 PM|SETI@home|[file_xfer] Finished download of file 05mr05aa.1632.21490.4826.3.200
5/21/2007 1:50:18 PM|SETI@home|[file_xfer] Throughput 6936 bytes/sec
5/21/2007 1:50:19 PM|SETI@home|Starting 05mr05aa.1632.21490.4826.3.200_3
5/21/2007 1:50:19 PM|SETI@home|Starting task 05mr05aa.1632.21490.4826.3.200_3 using setiathome_enhanced version 515

I just have to say, I think something is fishy, and I don't think it's all of our machines out here. I haven't seen any GWU's on any of my machines throughout this period. Just the project servers saying there's no work and constantly deferring communication unless I shut down BOINC.

I see a lot of participant commentary and conjecture on the boards, and helpful suggestions by volunteer testers, but I don't see anything from the actual SETI project team. Seriously, what do you guys have to definitively say about this problem; largely, we're out here burning electricity and pushing up gloabal warming for no good cause because of this. I think some Word From the Elder Gods would be a preferable encouragement to us all.

--R.

ID: 573186 · Report as offensive
mdpagel

Send message
Joined: 18 Sep 99
Posts: 53
Credit: 2,619,543
RAC: 0
United States
Message 573339 - Posted: 22 May 2007, 0:02:53 UTC - in response to Message 573009.  

I have tried the following thusfar:
redownload boinc and apply over the top of existing installation <--- still didn't download new workunits
Replace KWSN app with new version <--- still no download
Delete BOINC directory <-- failed to delete (windows complained)
Add/Remove programs on BOINC <-- didn't completely get rid of BOINC programs
Create new directory in BOINC directory and move all files into that directory <-- seemed to work
Re-install BOINC and re-attach <-- worked, did benchmarks, but still not downloading WUs. Futher, there is no client in my s@h directory (C:\\Program Files\\BOINC\\projects\\setiathome.berkeley.edu) - the directory is completely empty
Re-install optimized client <--- it's in the directory now at least;)

Currently statistics shows that I've completed no work ever, etc.

Boinc is a service installation on this computer.

I had gotten about 20 new WUs after the thumper issues, but after a while it stopped downloading additional units. I've gotten HTTP internal server errors, as well as Scheduler request failed: server returned nothing (no headers, no data), Access to reference site succeeded - project servers may be temporarily down. My "results" show that I have several workunits that seti thinks I've downloaded but haven't processed. I've never seen these WUs (before or after uninstall).



You must have created a new account rather than using your old account.



I don't believe so. I told it to use an existing account. I only typed in my password once (rather than in a create password and confirm password blank). I used the same e-mail address that this forum account and thus my RAC is under. Moreover, if I had created a new account, why would I not be getting workunits, despite requesting 17K seconds of work?
ID: 573339 · Report as offensive
Keith Huff (kshuff)
Avatar

Send message
Joined: 3 Apr 99
Posts: 10
Credit: 6,045,689
RAC: 0
United States
Message 573364 - Posted: 22 May 2007, 0:18:45 UTC - in response to Message 572357.  

I have a feeling this method only works in cases where your platform is officially supported. For those of you that MUST use an app_info.xml, I suspect you are out of luck for now.

This method should work fine for platforms windows_intelx86 and i686-pc-linux-gnu (at least, since I have these).


This sucks!



ID: 573364 · Report as offensive
Profile littlegreenmanfrommars
Volunteer tester
Avatar

Send message
Joined: 28 Jan 06
Posts: 1410
Credit: 934,158
RAC: 0
Australia
Message 573637 - Posted: 22 May 2007, 11:02:32 UTC - in response to Message 573172.  

Any idea how long it takes for the ghosts to disappear after doing the detach/reattach? Just curious - mine are still there as "client detached" I was able to do it on my second computer this morning so both should be clean of ghosts after the 'client detached'(ghosts) go away.


If they show as "client detached", they will remain in your results list until after the completion date. After that, they will disappear. As each WU reaches it's deadline, it will be removed, so the process may take a few weeks or days to clear all the aborted WUs caused by the detach.
ID: 573637 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14686
Credit: 200,643,578
RAC: 874
United Kingdom
Message 573643 - Posted: 22 May 2007, 11:13:14 UTC - in response to Message 573637.  

Any idea how long it takes for the ghosts to disappear after doing the detach/reattach? Just curious - mine are still there as "client detached" I was able to do it on my second computer this morning so both should be clean of ghosts after the 'client detached'(ghosts) go away.


If they show as "client detached", they will remain in your results list until after the completion date. After that, they will disappear. As each WU reaches it's deadline, it will be removed, so the process may take a few weeks or days to clear all the aborted WUs caused by the detach.

They may clear quicker or slower than this - the rate will vary from WU to WU.

Remember, it's actually the result which is the ghost. If your ghost has been matched with two stock crunchers, it'll get purged soon after the other two have reported in and been validated.

On the other hand, if your ghost has been matched with other ghosts, and they don't detach, then you'll have to wait for deadline - reissue - crunch - report - validate, with all the pitfalls that involves. We won't see the last of them until well into July.
ID: 573643 · Report as offensive
Profile littlegreenmanfrommars
Volunteer tester
Avatar

Send message
Joined: 28 Jan 06
Posts: 1410
Credit: 934,158
RAC: 0
Australia
Message 574049 - Posted: 23 May 2007, 0:06:42 UTC

Probs continuing here...
http internal server error, and the number of ghosts in results list is still rising.
ID: 574049 · Report as offensive
Profile Jason Safoutin
Volunteer tester
Avatar

Send message
Joined: 8 Sep 05
Posts: 1386
Credit: 200,389
RAC: 0
United States
Message 574224 - Posted: 23 May 2007, 5:22:31 UTC - in response to Message 574049.  

Probs continuing here...
http internal server error, and the number of ghosts in results list is still rising.


As long as you do this with optimized apps everytime you want new work, which is why I get a bunch then no new task it, change the app_info filename to whatever you want when you have no work in any pending stage...then restart BOINC, click update...get work...suspend...rename that file back to app_info...

Gotta do this everytime you want new work...so far the ghosts I got before doing that are expiring and I have yet to recieve anymore.

By doing the above, you also get a larger cache of work.
"By faith we understand that the universe was formed at God's command, so that what is seen was not made out of what was visible". Hebrews 11.3

ID: 574224 · Report as offensive
Profile littlegreenmanfrommars
Volunteer tester
Avatar

Send message
Joined: 28 Jan 06
Posts: 1410
Credit: 934,158
RAC: 0
Australia
Message 574869 - Posted: 24 May 2007, 14:26:35 UTC - in response to Message 574224.  
Last modified: 24 May 2007, 14:27:05 UTC

Probs continuing here...
http internal server error, and the number of ghosts in results list is still rising.


As long as you do this with optimized apps everytime you want new work, which is why I get a bunch then no new task it, change the app_info filename to whatever you want when you have no work in any pending stage...then restart BOINC, click update...get work...suspend...rename that file back to app_info...

Gotta do this everytime you want new work...so far the ghosts I got before doing that are expiring and I have yet to recieve anymore.

By doing the above, you also get a larger cache of work.


Hi Jason,

I have already renamed app_info.xml to oldapp_info.xml and got plenty of work. :) The optimised app still appeared to be doing the crunching, even though the stock app was downloaded along with the first WUs I received after renaming app_info. The http error occurred when uploading completed results and when updating, to report them.

Renaming the app_info.xml file back to it's original name after downloading WUs seemed to clear that problem, but the number of ghosts is still rising. :(

It would seem that the ghosts are appearing after clicking on "Update" to report completed WUs after uploading them.

I'm unsure whether I should therefore allow BOINC to do the reporting automatically, and whether that would stop the ghosts being generated?


ID: 574869 · Report as offensive
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 574872 - Posted: 24 May 2007, 14:33:56 UTC - in response to Message 574869.  
Last modified: 24 May 2007, 14:43:32 UTC

Hello, Please note that the 'Anonymous Platform' Mechanism in the server has been fixed, and you should return your app_info.xml and restart. This will provide a known starting point for further checks to see if you still have problems. If you have lots of ghosts we can work on releasing those later if need be or if you want to bother (you have to crunch cache dry and detach/reattach).

A simple Restart can fix much...

If you have lots of work in your cache you might want to set 'No New work' for now so we can start with seeing what errors uploads and reporting give you(if any).

[Different Jason :D]
"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.
ID: 574872 · Report as offensive
Previous · 1 . . . 9 · 10 · 11 · 12

Message boards : Number crunching : Ghost WU issue (and some talk about deadlines)


 
©2025 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.