| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Human Proteome Folding - Phase 2
Thread: Issue with HPFII on 64bit Quads running 64bit Vista. |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 72
|
|
| Author |
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Here's a sample of the 2 minutes failure HPF2's, as reported no longer suffered on the Vista Quad:
----------------------------------------<core_client_version>6.2.28</core_client_version> <![CDATA[ <message> Funzione non corretta. (0x1) - exit code 1 (0x1) </message> <stderr_txt> ERROR:: Exit at: .\dock_structure.cc line:401 </stderr_txt> ]]> The key error lines in italic Since the fix, 100% result score for this project. Also note an interesting post today by knreed http://www.worldcommunitygrid.org/forums/wcg/printpost?post=212271. He's discussing project level tracking of success rates and automatic torquing off devices that consistently fail on particular project, but only that project, meaning in the future you'd only be getting 1 HPF2 job over a X period to test if things have changed.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
123bob
Cruncher Joined: May 1, 2007 Post Count: 42 Status: Offline Project Badges: 
|
Sek,
On the machines I posted about, I'm not running any virus programs. They are stripped machines with only the regular vista stuff on them. I will still try looking into that possibility. Thx. Bob |
||
|
|
123bob
Cruncher Joined: May 1, 2007 Post Count: 42 Status: Offline Project Badges:
|
I guess there no way to edit my prior post, so I'll update here.
I took two machines that were showing the problem, Host names "El-Machino-3" and "BOBS-FARM-03". I moved them to a profile that has HPF2 as the primary project. On El-Machino-03, I completely turned off windows firewall. It will take a couple of days to work them through the queue, but we'll see how they run. I'll report back. It does appear that our status logs are identical. (Except for the language difference of course.) I hope this turns something up. Regards, Bob |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hello 123bob,
Edit is a pen icon in the upper left side of your old post. Click on the pen and you can edit your post. Lawrence |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Windows Defender too.
----------------------------------------
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
123bob
Cruncher Joined: May 1, 2007 Post Count: 42 Status: Offline Project Badges:
|
OK, I have some run results to report.
I have el-machino-3 and el-machino-4 running HPF2. Both machines are Vista ultimate 64 bit. el-m-3 has everything in the security suite turned off. No firewall, virus protection, win defender, etc. It's completely naked. It is a Q6600 G0 stepping quad. 4 gig RAM, and a very mild overclock. El-m-4 has only the vista security stuff running, win firewall, defender, etc. El-m-4 has no virus program running, of any kind, at the moment. It is a dual Xeon (8 cores) with 8 gig of RAM, running completely stock clocks. (It is my "main rig".) So far, El-m-3 reports 6 errored WUs with the same result as the above posted result logs. It has 5 apparently successful pending validation WUs. El-m-4 reports 7 errored WUs with the same logs as above, and it has 6 WUs in the pending val queue. I'm seeing no difference between the fail rate between the two machines so far. The fail rate is above 50%!!.......  I'll let them run for a bit to gather more of a sample.I will also throw an old faithful XP pro 32 bit on HPF2 (El-Machino-2) to use as a control. I would anticipate it to run clean..... Regards, Bob |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
 Thanks, 123bob. I am completely puzzled by this, but until it is solved I will recommend not running HPF2 on Vista 64 machines that give error results like this. Lawrence |
||
|
|
123bob
Cruncher Joined: May 1, 2007 Post Count: 42 Status: Offline Project Badges:
|
OK, I came back to the test after a week on the road. The machines, El-Machino-3 and El-Machino-4, running Vista 64 bit are full of pages of errors as reported above. The single machine, El-Machino-2, running XP 32 bit pro is completely clean.
So, we can conclude this is isolated to a Vista issue. I don't know if others are seeing this on Vista 32 bit installations. That would be useful to know. Please add to this thread if you have a Vista 32 bit install running HPF2. Sek reported that his machine (on Vista 64 bit?) is running clean. I'd like to find out why..... Regards, Bob |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
32 bit, and the last 10 started all went down the drink again out of the 45 i see for that machine in the RS list. Some whilst other HPF2 were still running that completed correctly. (BOINCview logs the time in the "completed" window).
----------------------------------------Is it Quad+Vista? Is it 6.2+Quad+Vista? Is it 6.2+Quad+Vista+FAAH running on other cores? There was a series of 35 that did fine after a series of 6 that died quick. Curiously, the last 10 fails occurred after the last reboot. And all with the log line: <core_client_version>6.2.28</core_client_version> <![CDATA[ <message> Funzione non corretta. (0x1) - exit code 1 (0x1) </message> <stderr_txt> ERROR:: Exit at: .\dock_structure.cc line:401 </stderr_txt> Something is conflicting, but what? For sure, I can't run RICE on 6.2.+DUO+XP,32bit. Runs fine on 5.10.45... so that might be my next test... reverting to 5.10.45 on Vista, though I see a log in previous post with that client and crashes happening. Maybe the BOINC 6 API files to compile the sciences?
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I've got a Vista 32 machine generating plenty of these errors (although plenty also work fine.)
|
||
|
|
|