Hi all. On all running projects lately I've started to see a 'Couldn't delete file...' error in the message log right after an upload of data after a wu is done. I've also been starting to get the same errors lately as on these two threads, Disconnected and lost connection to local host. I have not updated Win2k or IE6.0, nor have any changes been made to ZoneAlarm. Also, on top of this, I'm running an older core client of BOINC, v5.2.13.

Here's info on the computer, which is on Win2k:

12/17/2006 10:56:44 PM||Starting BOINC client version 5.2.13 for windows_intelx86
12/17/2006 10:56:44 PM||libcurl/7.14.0 OpenSSL/0.9.8 zlib/1.2.3
12/17/2006 10:56:44 PM||Data directory: C:\\Program Files\\BOINC
12/17/2006 10:56:51 PM||Processor: 1 GenuineIntel Intel(R) Pentium(R) 4 Mobile CPU 1.60GHz
12/17/2006 10:56:51 PM||Memory: 766.98 MB physical, 1.08 GB virtual
12/17/2006 10:56:51 PM||Disk: 9.76 GB total, 1.50 GB free

Running Einstein, SETI, rosetta, Predictor, lhcathome, Pirates, ralph, uFluids, Leiden, and Nano.

Started noticing the issues when these errors were showing up in the message log on all projects after an upload of data:
12/18/2006 12:00:46 AM|uFluids|Couldn't delete file projects/www.ufluids.net/axi_linear_incomp1_0.5_1_72_0_0_1_-1.29999999999999_400_-1.29999999999999_0.01_0_0

I performed the repair installation and it seemed to help for only a little, maybe half a day, then the 'Couldn't delete file...' errors started showing up again.

I researched back through my message log data (I save it everytime I shut down the laptop to go to a different location) and found the first couple of instances of the error and what seems to might have caused it:

Edit, added the following:
12/08/2006 07:43:51 PM|Einstein@Home|Started upload of h1_1323.5_S5R1__824_S5R1a_0_0
12/08/2006 07:43:51 PM|Einstein@Home|Fetching master file
12/08/2006 07:43:52 PM|Einstein@Home|Error on file upload: Maintenance underway: file uploads are temporarily disabled.
12/08/2006 07:43:52 PM|Einstein@Home|Temporarily failed upload of h1_1323.5_S5R1__824_S5R1a_0_0: transient upload error
12/08/2006 07:43:52 PM|Einstein@Home|Backing off 3 hours, 18 minutes, and 12 seconds on upload of file h1_1323.5_S5R1__824_S5R1a_0_0
12/08/2006 07:43:56 PM|Einstein@Home|Master file download succeeded
12/08/2006 07:44:02 PM|Einstein@Home|Sending scheduler request to http://einstein.phys.uwm.edu/EinsteinAtHome_cgi/cgi
12/08/2006 07:44:02 PM|Einstein@Home|Reason: Requested by user
12/08/2006 07:44:02 PM|Einstein@Home|Note: not requesting new work or reporting results
12/08/2006 07:44:07 PM|Einstein@Home|Scheduler request to http://einstein.phys.uwm.edu/EinsteinAtHome_cgi/cgi succeeded
12/08/2006 07:44:07 PM|Einstein@Home|Message from server: Project is temporarily shut down for maintenance
12/08/2006 07:44:07 PM|Einstein@Home|Project is down

12/08/2006 11:02:06 PM|Einstein@Home|Started upload of h1_1323.5_S5R1__824_S5R1a_0_0
12/08/2006 11:02:15 PM|Einstein@Home|Finished upload of h1_1323.5_S5R1__824_S5R1a_0_0
12/08/2006 11:02:15 PM|Einstein@Home|Throughput 22420 bytes/sec

12/09/2006 01:26:18 AM|Einstein@Home|Sending scheduler request to http://einstein.phys.uwm.edu/EinsteinAtHome_cgi/cgi
12/09/2006 01:26:18 AM|Einstein@Home|Reason: To report results
12/09/2006 01:26:18 AM|Einstein@Home|Reporting 1 results
12/09/2006 01:26:23 AM|Einstein@Home|Scheduler request to http://einstein.phys.uwm.edu/EinsteinAtHome_cgi/cgi succeeded

12/10/2006 12:22:04 AM|Einstein@Home|Sending scheduler request to http://einstein.phys.uwm.edu/EinsteinAtHome_cgi/cgi
12/10/2006 12:22:04 AM|Einstein@Home|Reason: Requested by user
12/10/2006 12:22:04 AM|Einstein@Home|Requesting 8640 seconds of new work
12/10/2006 12:22:09 AM|Einstein@Home|Scheduler request to http://einstein.phys.uwm.edu/EinsteinAtHome_cgi/cgi succeeded
12/10/2006 12:22:11 AM|Einstein@Home|Started download of grid_1330_h_T04_S5R1.dat
12/10/2006 12:22:15 AM||request_reschedule_cpus: project op
12/10/2006 12:22:16 AM|Einstein@Home|Finished download of grid_1330_h_T04_S5R1.dat
12/10/2006 12:22:16 AM|Einstein@Home|Throughput 165155 bytes/sec
12/10/2006 12:22:17 AM||request_reschedule_cpus: files downloaded

12/10/2006 12:34:36 AM|Einstein@Home|Starting result h1_1323.5_S5R1__660_S5R1a_1 using einstein_S5R1 version 424
12/10/2006 12:34:36 AM|Einstein@Home|Sending scheduler request to http://einstein.phys.uwm.edu/EinsteinAtHome_cgi/cgi
12/10/2006 12:34:36 AM|Einstein@Home|Reason: Requested by user
12/10/2006 12:34:36 AM|Einstein@Home|Note: not requesting new work or reporting results
12/10/2006 12:34:41 AM|Einstein@Home|Scheduler request to http://einstein.phys.uwm.edu/EinsteinAtHome_cgi/cgi succeeded

Finish Edit. The file upload error was when Einstein's file servers went down. Maybe they caused an error in one/some of the files downloaded to BOINC and caused a bug. Right after the Unrecoverable error below you can see the first instance of the 'Couldn't delete file..' error. Just a guess that the corrupt Einstein wu caused it.

12/10/2006 01:14:20 AM|Einstein@Home|Unrecoverable error for result h1_1323.5_S5R1__660_S5R1a_1 (The environment is incorrect. (0xa) - exit code 10 (0xa))
12/10/2006 01:14:20 AM||request_reschedule_cpus: process exited
12/10/2006 01:14:20 AM|Einstein@Home|Computation for result h1_1323.5_S5R1__660_S5R1a_1 finished
12/10/2006 01:14:21 AM|uFluids|Resuming result bannulus4_35_1.021_0.002_1.1_0_0_0_1 using evolver version 410
12/10/2006 01:14:23 AM|Einstein@Home|Started upload of h1_1323.5_S5R1__660_S5R1a_1_0
12/10/2006 01:16:39 AM|lhcathome|Sending scheduler request to http://lhcathome.cern.ch/lhcathome_cgi/cgi
12/10/2006 01:16:39 AM|lhcathome|Reason: To fetch work
12/10/2006 01:16:39 AM|lhcathome|Requesting 8640 seconds of new work
12/10/2006 01:16:44 AM|lhcathome|Scheduler request to http://lhcathome.cern.ch/lhcathome_cgi/cgi succeeded
12/10/2006 01:16:44 AM|lhcathome|No work from project
12/10/2006 01:17:17 AM|Einstein@Home|Finished upload of h1_1323.5_S5R1__660_S5R1a_1_0
12/10/2006 01:17:17 AM|Einstein@Home|Throughput 30777 bytes/sec
12/10/2006 01:17:19 AM|Einstein@Home|Couldn't delete file projects/einstein.phys.uwm.edu/h1_1323.5_S5R1__660_S5R1a_1_0
12/10/2006 01:25:00 AM|NanoHive@Home|Sending scheduler request to http://www.nanohive-1.org/atHome_cgi/cgi
12/10/2006 01:25:00 AM|NanoHive@Home|Reason: Requested by user
12/10/2006 01:25:00 AM|NanoHive@Home|Requesting 8640 seconds of new work
12/10/2006 01:25:05 AM|NanoHive@Home|Scheduler request to http://www.nanohive-1.org/atHome_cgi/cgi succeeded
12/10/2006 01:25:05 AM|NanoHive@Home|Message from server: Project is temporarily shut down for maintenance
12/10/2006 01:25:05 AM|NanoHive@Home|Project is down
12/10/2006 01:27:54 AM||request_reschedule_cpus: process exited
12/10/2006 01:27:54 AM|uFluids|Computation for result bannulus4_35_1.021_0.002_1.1_0_0_0_1 finished
12/10/2006 01:27:54 AM|Leiden Classical|Resuming result wu_78596477_1165394153_2249_2 using classical version 535
12/10/2006 01:27:56 AM|uFluids|Started upload of bannulus4_35_1.021_0.002_1.1_0_0_0_1_0
12/10/2006 01:28:02 AM|uFluids|Finished upload of bannulus4_35_1.021_0.002_1.1_0_0_0_1_0
12/10/2006 01:28:02 AM|uFluids|Throughput 724 bytes/sec
12/10/2006 01:28:04 AM|uFluids|Couldn't delete file projects/www.ufluids.net/bannulus4_35_1.021_0.002_1.1_0_0_0_1_0

This weekend I started seeing a error message window popup and have to restart BOINC, I think it was an Unhandled exception error (sorry, I didn't copy it).

On several instances this weekend I have also found BOINC to have lost communication and was disconnected. Tried to reconnect using File-Select Computer and entering localhost, but communication would fail. The only way to reconnect was to Exit BOINC and restart it.

Hope this helps since this is happening on a different client and a different OS than on the other two threads. Good luck and Happy Holidays!

---

This weekend I started seeing a error message window popup and have to restart BOINC, I think it was an Unhandled exception error (sorry, I didn't copy it).

If it was an Unhandled Exception, it was recorded in your Event Viewer.

So please check Start->Settings->Control Panel->Administrative Tools->Event Viewer. Check your Application and System logs.

---

Hi Jord. Thanks for the reply. Sorry, I was wrong. It was an Application Error popup. Here's the details from the Event Viewer System Log:

Event Type: Information
Event Source: Application Popup
Event Category: None
Event ID: 26
Date: 12/17/2006
Time: 07:10:59 PM
User: N/A
Computer: ***************
Description:
Application popup: evolver_4.10_windows_intelx86.exe - Application Error : The application failed to initialize properly (0xc0000044). Click on OK to terminate the application.

The the lost communication issues were new this weekend.

Looking back at the history I found three other popup errors related to a BOINC app as follows:

Event Type: Information
Event Source: Application Popup
Event Category: None
Event ID: 26
Date: 12/13/2006
Time: 09:11:21 AM
User: N/A
Computer: ***************
Description:
Application popup: classical_5.35_windows_intelx86.exe - Application Error : The application failed to initialize properly (0xc0000044). Click on OK to terminate the application.

Event Type: Information
Event Source: Application Popup
Event Category: None
Event ID: 26
Date: 12/13/2006
Time: 09:16:18 AM
User: N/A
Computer: ***************
Description:
Application popup: evolver_4.10_windows_intelx86.exe - Application Error : The application failed to initialize properly (0xc0000142). Click on OK to terminate the application.

Event Type: Information
Event Source: Application Popup
Event Category: None
Event ID: 26
Date: 12/13/2006
Time: 09:18:13 AM
User: N/A
Computer: ***************
Description:
Application popup: evolver_4.10_windows_intelx86.exe - Application Error : The application failed to initialize properly (0xc0000142). Click on OK to terminate the application.

Hope this helps. Thanks again for the help!

Hi all. Still having the 'Couldn't delete file...' error messages. Haven't had any other errors though. Any help would be great. Thanks!

Bump. Upgraded to v5.4.11, but still having the issue. Any help would be great. Thanks in advance and Happy New Year!

Is it possible that Boinc was originally installed with a different user account (such as administrator) to the one that you normally use?

---

Thanks for the response Mike. Nope, it was installed on the same user (the only one) on the unit. I'm thinking about doing the complete reinstall where you delete the entire directory after uninstalling. I'm trying to find good info on this so as not to screw up the projects, although it might be a good idea to have those reinstall themselves too, but I'm not sure how to do that.

Thanks again!

Thanks for the response Mike. Nope, it was installed on the same user (the only one) on the unit. I'm thinking about doing the complete reinstall where you delete the entire directory after uninstalling. I'm trying to find good info on this so as not to screw up the projects, although it might be a good idea to have those reinstall themselves too, but I'm not sure how to do that.

I believe Task Manager will show you the owner of the BOINC and project processes -- i.e. the account under which they were installed -- in case something got scrambled during or after installation.

As for exiting projects gracefully, in case it comes to that, just set each one to No New Tasks from the Projects pane. When all the queued work is complete, if there are still tasks showing as "Ready to report" you can clear them by Updating the corresponding project one last time. Then Detach your host from each project.

After a clean install, attach to each project again as an existing user, providing the same e-mail address & password you used before. The only issue I can foresee is that your host may appear twice -- in old and new versions -- in your accounts. If this (harmless) situation persists for more than a few days or just annoys you, you should be able to merge the entries, or at least delete the old one.

To safely clear all programs and data from a single project folder, run out the cache as above, then Reset the project after the final update. This should give you a clean slate, automatically downloading a fresh version of the app(s) you need for the work your host is assigned, without affecting any other projects. IIANM just detaching & reattaching doesn't remove any files: if BOINC finds data for that project on reattaching, it picks up exactly where it left off.

---

Thanks for the reply Odysseus. I looked in TM and couldn't find the owner/account for the task running. I checked to see if I could add the column (View-Select Columns...) but there was none for owner/account. Maybe it depends on your OS. I am on Win2k.

I have since did a 'slightly drastic' approach to try and fix the issue, but it didn't work. I set each project to 'No New Tasks' and let them run out, upload, and report. I then Uninstalled BOINC and then deleted everything remaining including the subdirectories except for the account*.xlm files, although I couldn't delete some of the files (had the error - 'Couldn't delete [filename]: There has been a sharing violation. The source or destination file may be in use.). I also found that Slot 1 had instances of multiple projects in it.

Next I rebooted to Safe Mode and deleted the files I could not under my user account, did a disk scan, then defrag'ed the drive. I then rebooted to my user account and reinstalled BOINC.

I then downloaded a couple of project wus, Einstein and uFluids, and set the projects back to 'No New Work'. The Eistein wu ran a couple of times, switching in and out with the UF wus, but failed on one of the restarts (I have not been able to have an Einstein wu complete without failure since the crash in early Dec). The UF@H wus completed ok and uploaded. The failed E@H uploaded. When each project uploaded the files, I got the 'Couldn't delete file...' error, then got it again when the wus reported.

I have since left E@H at No New Work and I'm letting UF@H, Pirates, and Leiden run. These projects are running fine, but still cannot delete files when uploading completed work and reporting work.

I would've thought that uninstalling and deleting, then reinstalling would have cleared the issue, but no luck. I've have found out that on Win2k if you have the drive as FAT32 (which is what my drive is) and not NTFS you cannot see the permissions on the file properties, so maybe there is another way to check this but I don't know it.

Thanks again for the help and have a great week!

Bump.

Still having the same issue. Please, can someone help with this. It is getting rather time consuming rebooting to Safe Mode and deleting the files that BOINC can't. I'm almost to the point where I am going to stop BOINC (I would like to continue to donate time to projects).

Thanks again!!!

Brian.

Not that I have any idea of what might be going wrong...

But you can try running the latest alpha client. I don't know what could be strange about the current release version in combo with your OS and hardware.

I've heard of stranger things being fixed by trying the alpha client.

That's really my only thought aside from seeing if your computer will pass MemTest or Prime95 (see this page in the Wiki.

---

Kathryn :o)

Hi Kat. Thanks again. Since I had the same issue with v5.2.13 as well as v5.4.11, I don't think it's a release issue.

I really think that when Einstein crashed BOINC and the OS back in early Dec it screwed something up, as this issue came up right after the crash; this was when they started having their issues. Upgrading the client didn't fix it, and uninstalling and deleting everything then reloading didn't fix it either.

Maybe it's a registry thing or a permissions thing like you suggested before. I'm on Win2k and I've found out that you can't check the permissions of the files unless its NTFS (I'm FAT32). Do you know if there is a workaround to check the permissions?

I'm going to try Memtest86+ tonight and let it run overnight. We'll see what happens.

Thanks again Kat for the reply and have a great week!!

Hi Kat. Thanks again. Since I had the same issue with v5.2.13 as well as v5.4.11, I don't think it's a release issue.

But on 19 Jan 2007 a change was made to the filesystem operations, and they are supposed to fix this. That would be version 5.8.4 (although note it has some other problems; stay tuned for next release).