Getting tired of this error

Message boards : Number crunching : Getting tired of this error

To post messages, you must log in.

AuthorMessage
Usuario1_S

Send message
Joined: 24 Mar 14
Posts: 92
Credit: 3,059,705
RAC: 0
Message 77813 - Posted: 9 Jan 2015, 11:13:02 UTC

09/01/2015 06:31:25 | rosetta@home | [error] Signature verification failed for minirosetta_database_3d2618f.zip
09/01/2015 06:31:32 | rosetta@home | Computation for task YIHY_cst__t000___krypton_SAVE_ALL_OUT_03_09_233081_3802_0 finished
09/01/2015 06:31:32 | rosetta@home | Output file YIHY_cst__t000___krypton_SAVE_ALL_OUT_03_09_233081_3802_0_0 for task YIHY_cst__t000___krypton_SAVE_ALL_OUT_03_09_233081_3802_0 absent
09/01/2015 06:31:37 | rosetta@home | Computation for task Azurin_2ring_3_320_fold_SAVE_ALL_OUT_228261_277_0 finished
09/01/2015 06:31:37 | rosetta@home | Output file Azurin_2ring_3_320_fold_SAVE_ALL_OUT_228261_277_0_0 for task Azurin_2ring_3_320_fold_SAVE_ALL_OUT_228261_277_0 absent

Screenshots:
http://sd.uploads.im/uSXUT.jpg
http://sd.uploads.im/IleG7.jpg

I reinstalled Win7 64-bit, After 2 days of a new BOINC installation, first it gives a error message for 1 of the WUs, 1 day later the master database is corrupt and all WUs give error, my hard drive is OK, I have no hardware errors reported by Windows Error Reporting, this has happened to me non-stop since about 1 month or so, if this doesn't get solve in the next 2 weeks, I'm out of here.
ID: 77813 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Usuario1_S

Send message
Joined: 24 Mar 14
Posts: 92
Credit: 3,059,705
RAC: 0
Message 77814 - Posted: 9 Jan 2015, 12:57:16 UTC - in response to Message 77813.  

PS: your forum sucks, you could move this forum to forumsland.com it's free and has nice looking skins. Or at least use https://www.phpbb.com but do something this way of having an infinite number of posts in a thread is just highly inefficient and archaic.
ID: 77814 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Polian
Avatar

Send message
Joined: 21 Sep 05
Posts: 152
Credit: 10,141,266
RAC: 0
Message 77816 - Posted: 10 Jan 2015, 0:27:34 UTC

The screenshots given do not provide any useful troubleshooting information. From the log it appears that the database file is failing a hash comparison check, i.e. the contents of the file as stored on your computer doesn't match known good values. Is the download getting interrupted and the file truncated? Are you using any sort of data compression or proxy server (or any other non-standard configuration) on your internet connection? Sorry, but there just isn't widespread reporting of your problem, which suggests something is wrong on your end.

and PS the forum is integrated with the server software
ID: 77816 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Usuario1_S

Send message
Joined: 24 Mar 14
Posts: 92
Credit: 3,059,705
RAC: 0
Message 77817 - Posted: 10 Jan 2015, 3:59:52 UTC - in response to Message 77816.  

No compression nor proxy, no corrupted download, Internet is OK, As I described it get's used 2 days normally, I start and exit BOINC several times a day, so unless the check is selective for the part o0f the research Rosetta it's going to do this makes no sense, normally I would just let it auto-reset the project: deleting all files, downloading another copy of the database and the WUs, but it's just going to happen again and again, I let it go for 5 times to see if it was recurring and if it was something special each time, but no, all the WUs seem to be different on each reset, it can't be my Hard Drive, windows will report a SMART error, BIOS will report the same at boot time, so this looks like Rosetta executables are wrong, because I have been crunching normally for month and then this.

Right now I have suspended Rosetta and running WCG FightAIDS@Home to check functionality for 3 days.
ID: 77817 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
sgaboinc

Send message
Joined: 2 Apr 14
Posts: 282
Credit: 208,966
RAC: 0
Message 77819 - Posted: 12 Jan 2015, 16:40:24 UTC

i'm not too sure if u try doing a 'reset project' if that may help.

as one of the messages appear to be:
09/01/2015 06:31:25 | rosetta@home | [error] Signature verification failed for minirosetta_database_3d2618f.zip

this seem to be a common file that i'd guess works like a 'reference dictionary' when rosetta runs.

i'd think if you do a 'reset project' it may delete the existing files & download a fresh copy rosetta executable as well as this zip file, hope it helps
ID: 77819 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Snags

Send message
Joined: 22 Feb 07
Posts: 198
Credit: 2,888,320
RAC: 0
Message 77820 - Posted: 12 Jan 2015, 22:38:49 UTC

A spot check through your task lists show other folks have successfully completed the workunits which does suggest the problem originates at your end. I don't run Windows so I can't be of much help but I did notice something while scrolling through the task list which may trigger a helpful tip from someone else.

A bunch of tasks were marked "client detached" on January 4th. I assume this is when you reset the project. Of the new tasks downloaded on the 4th all were completed and returned successfully on the 6th and 7th of January. Of the new tasks assigned on the 6th most were returned and completed successfully on the 7th and 8th. On the 9th only two workunits were returned successfully. The rest were completed successfully by your machine but were marked "client error" with this at the bottom of the stdrr out:

======================================================
DONE :: 1460 starting structures 21586.7 cpu seconds
This process generated 1460 decoys from 1460 attempts
======================================================
BOINC :: WS_max 4.82271e+008

BOINC :: Watchdog shutting down...
BOINC :: BOINC support services shutting down cleanly ...
called boinc_finish

</stderr_txt><message>
app_version download error: couldn't get input files:
<file_xfer_error>
<file_name>minirosetta_database_3d2618f.zip</file_name>
<error_code>-120 (RSA key check failed for file)</error_code>
<error_message>signature verification failed</error_message>
</file_xfer_error>

</message>

(These talks were awarded credits, by the way, as you appear to have been able to send back usable information).

Of the tasks assigned:
6 Jan 2015 7:41:50 UTC or earlier (after the rest on the 4th) completed and reported just fine.
6 Jan 2015 7:54:04 UTC all completed models but reported with the signature verification error message.
7 Jan 2015 6:05:08 UTC two were returned in the same fashion but the rest, and all subsequently received tasks were reported with no work completed (zero CPU time used, no models run) along with the same error message.


Could this be triggered by firewall/security issues?

HTH

Snags



ID: 77820 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Timo
Avatar

Send message
Joined: 9 Jan 12
Posts: 185
Credit: 45,649,459
RAC: 0
Message 77821 - Posted: 13 Jan 2015, 1:28:54 UTC
Last modified: 13 Jan 2015, 1:30:09 UTC

Most all of the tasks that are ending in error on your machine are finishing without any issue on other people's machines so it is definitely something local to your machine. It sounds like somehow the database file is getting corrupted - it is possible that it may be some firewall, security software, antivirus, etc. or even a driver issue.

Because checksums are failing, it would also be prudent to check your timezone/date/time settings (some checksum and security certificates are time-dependent and may not work correctly if the timezone is not set properly) and also check any proxy settings.

I am fairly confident it is not an issue with the rosetta software as if it were the issues you're showing would be much more widely reported.
ID: 77821 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Usuario1_S

Send message
Joined: 24 Mar 14
Posts: 92
Credit: 3,059,705
RAC: 0
Message 77847 - Posted: 22 Jan 2015, 15:32:27 UTC - in response to Message 77821.  

I've been doing some testing:

- To discard is a physical part of my hard drive, I have defragmented with PerfectDisk 13 Build 842, then

- Ran FightAIDS@Home on WCG, 4 days, no problems.

- About to finish climateprediction.net 4 days, no problems so far.

Tomorrow I'll try Rosetta again for 4 days and see how ti goes.
ID: 77847 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Mod.Sense
Volunteer moderator

Send message
Joined: 22 Aug 06
Posts: 4018
Credit: 0
RAC: 0
Message 77848 - Posted: 22 Jan 2015, 17:03:45 UTC

If you are currently detached from R@h, go in to your BOINC data directory (location shown in event log as BOINC Manager first starts) in to the rosetta subdirectory, and remove any files you see there. I couldn't recall if the "sticky" files get removed when you detach.

example windows path: C:ProgramDataBOINCprojectsboinc.bakerlab.org_rosetta
Rosetta Moderator: Mod.Sense
ID: 77848 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Usuario1_S

Send message
Joined: 24 Mar 14
Posts: 92
Credit: 3,059,705
RAC: 0
Message 77854 - Posted: 25 Jan 2015, 6:28:02 UTC - in response to Message 77848.  

I have done that several times, but isn't needed, as I wrote Rosettas@Home resets the project, so it deletes everything except login data, and done it 3 times and it's the same error.

I got an computation error at the end of a 196 hour WU on climateprediction.net:
24/01/2015 18:45:22 | climateprediction.net | Output file hadam3p_anz_a2l8_2013_1_009461659_0_12.zip for task hadam3p_anz_a2l8_2013_1_009461659_0 absent

how can this be? it's like something deleted it, maybe is a virus/spyware or a bug? I made a chkdsk on all my units and no errors, maybe the project didn't saved the data properly. Anyway, climateprediction.net didn't uploaded the 4 WUs in queue and just discarded them.

Now I'm back to Rosetta@Home and so far no problems.
ID: 77854 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Usuario1_S

Send message
Joined: 24 Mar 14
Posts: 92
Credit: 3,059,705
RAC: 0
Message 77859 - Posted: 27 Jan 2015, 10:05:32 UTC - in response to Message 77854.  

26/01/2015 20:19:05 | rosetta@home | [error] Signature verification failed for minirosetta_database_3d2618f.zip

Is there a Signature verification software I can download to test this as a standalone on my PC? because this would indicate a RAM failure, which is not probable because of me extensive test a while ago, I even reinstalled my Win 7 64-bit, maybe there is a bug on the Rosetta@Home software, this bug would be invisible if you have a high speed Internet like 10 Mbps or higher, and you have a BOINC to access the Internet at all times: Rosetta@Home would reset the project, download more files fast and keep processing WUs, so unless you check the log window a normal user would just see it as nothing is wrong.

But I had an error on the climate project but that could be a bug too, I have never had an error on WCG, running many sub-projects in it, even big ones like Clean Energy 2 which used a lot of RAM and generated huge results and I was able to upload them all fine.

SO unless I can run the verifier on my PC I'm out, I'll wait a week for an answer after that I'm moving to another BOINC project.
ID: 77859 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Snags

Send message
Joined: 22 Feb 07
Posts: 198
Credit: 2,888,320
RAC: 0
Message 77860 - Posted: 27 Jan 2015, 17:36:45 UTC

Could you clarify something? When you write that rosetta@home resets the project do you mean that all the files get deleted without any manual intervention on your part? I had assumed you meant you had clicked the "reset project" command from within the BOINC Manager. If in fact the files are being deleted without your having clicked that button (or the button below labeled "remove") then I suggest you refocus your troubleshooting to the activities of your security software. The pattern in your tasks list (first successful completions followed by unsuccessful completions, all of tasks downloaded at the same time) supports the hypothesis that security software running on an automated schedule is deleting files.

Timo also suggested that you check your timezone/date/time settings. I vaguely recall some old, possibly ancient issue with the Windows system clock which caused it to reset every day causing (I believe) the checksum error. As I mentioned earlier, I don't run Windows machines and I have no idea if it is still an issue.

Best,
Snags
ID: 77860 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Usuario1_S

Send message
Joined: 24 Mar 14
Posts: 92
Credit: 3,059,705
RAC: 0
Message 77863 - Posted: 28 Jan 2015, 7:57:36 UTC - in response to Message 77860.  

I don't know why people have trouble understanding, I clearly wrote on posts before that I have manually gone to the BOINC Data Directory and deleted all files as well let Roseeta@Home reset itself, many times. I tried a couple of times to reset manually Rosetta@Home by BOINC command and it's irrelevant, the 3 reset types make nothing to avoid the same errors, maybe it's my PC by everything indicates that it's not.

The reason I caught this error is because I suspend network activity for BOINC within the program because I uploads a lot so I need the bandwidth which isn't much, 25 KB/s, I allow the BOINC to use the network once a day approx. so any error is caught.

I have tested my RAM again with Microsoft Memory Diagnostics: 2 pass extended test, default CPU Cache settings; no errors.

Long ago I made a BIOS upgrade, and I remembered that since then I got small errors so I reverted to the original BIOS that worked fine, what I have left to do is reinstall my Win7 64-bit again.

I use Avast! and COMODO Firewall with Defense+, both are disabled except Comodo Firewall, I only activcate themwhen I need to introduce files to my PC.

But I'll check again all that you have mentioned just be sure.
ID: 77863 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Usuario1_S

Send message
Joined: 24 Mar 14
Posts: 92
Credit: 3,059,705
RAC: 0
Message 77864 - Posted: 28 Jan 2015, 9:51:22 UTC - in response to Message 77863.  

My time zone it was set right when I registered, no issue there, my clock is on time.

I checked Windows Defender at it wasn't stopped properly at the seems, now it should, maybe was this, we'll see, I'll try the climate project again and then Rosetta and if both fails I'll reinstall WIn7 Pro 64-bit
ID: 77864 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Mod.Sense
Volunteer moderator

Send message
Joined: 22 Aug 06
Posts: 4018
Credit: 0
RAC: 0
Message 77867 - Posted: 28 Jan 2015, 15:52:42 UTC

Others have had issues with Comodo as well. See details here.
Rosetta Moderator: Mod.Sense
ID: 77867 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Usuario1_S

Send message
Joined: 24 Mar 14
Posts: 92
Credit: 3,059,705
RAC: 0
Message 77883 - Posted: 1 Feb 2015, 16:40:16 UTC - in response to Message 77867.  

Others have had issues with Comodo as well. See details here.


I think I got it, yes, probably was Comodo Defense+ related, not the Firewall. The Climate prediction Project gave another error, I checked the event log and again the files were prevented from being created, I thought it was a blocked executable and yhe only possibility was Comodo Firewall/D+ so I checked the Firewall and as before no blocking there and wouldn't have prevented for any file to be created, but I saw 44 Unrecognized Files on the Overview, I click on it and took me to the window and yes there they were, some BOINC executables and even Excel, this is under File Rating, on its Settings Enable Cloud Lookup was enabled, I disabled it, then I put BOINC Data folder which I have it on a custom path and the installed BOINC folder under Trusted Files. Also created on Comodo a File Group called 'Trusted' and put the mentioned folders there and add it to Defense+ as Allowed Applications, in case I ever enable it.

I'll run the climate project for a few more days to check files aren't being blocked and then try Rosetta again.

Comodo Firewall is a great Security System you just need to know it, and if I hadn't forget it to check all settings, I probably wouldn't be having trouble with it, so it was me not the software. I read the few last post related to Comodo on the thread you pointed out and the guy obviously didn't know more than the basics, so "I tried all I knew about it" wasn't enough.
ID: 77883 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Usuario1_S

Send message
Joined: 24 Mar 14
Posts: 92
Credit: 3,059,705
RAC: 0
Message 77941 - Posted: 14 Feb 2015, 9:02:26 UTC - in response to Message 77883.  
Last modified: 14 Feb 2015, 9:05:01 UTC

The climate Project, ran smoothly except when my PC hanged and I made a cold reboot (reset), pretty sure it was a bug on the climate project. Happened again, after a day, looks like is related in part to how you exit BOINC, if you just restart Windows without exiting BOINC, next it's run, will give this signature error.

I checked Comodo and no .exe have been marked as 'unrecognized', pretty sure is a bug, and since people of this project don't bother to try to fix it, I'm moving on, too bad, I was doing this for my uncle but sucha sloppy project I don't think it's going to yield something soon or with good integrity. Anyway I'll deliver the few last results when they're finished.

Bye.
ID: 77941 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Usuario1_S

Send message
Joined: 24 Mar 14
Posts: 92
Credit: 3,059,705
RAC: 0
Message 77947 - Posted: 15 Feb 2015, 13:31:31 UTC - in response to Message 77941.  

I made a chodsk again after another hang and discovered a small failure in the file system, free space assigned as used, and now I have no errors for 2 days, if things go like this I'll stay.
ID: 77947 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Usuario1_S

Send message
Joined: 24 Mar 14
Posts: 92
Credit: 3,059,705
RAC: 0
Message 77992 - Posted: 4 Mar 2015, 8:48:34 UTC - in response to Message 77947.  

You were right, it was my side, I didn't clean my mother board in years so a small patches of mold were growing on the back of it, I cleaned it up and made a general cleaning too, and every thing is OK now.
ID: 77992 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote

Message boards : Number crunching : Getting tired of this error



©2024 University of Washington
https://www.bakerlab.org