For the record:

My Tasks for User Account insists that my computer has received the following wus:

https://boinc.bakerlab.org/rosetta/workunit.php?wuid=149489497
https://boinc.bakerlab.org/rosetta/workunit.php?wuid=149488655
https://boinc.bakerlab.org/rosetta/workunit.php?wuid=149486989
https://boinc.bakerlab.org/rosetta/workunit.php?wuid=149488150

The log file of my computer says no wus at all have been received in the period indicated (morning May 15th at 9:11 and 9:17 UTC). No such wus have been completed or are waiting on line, so they seem to have vanished somewhere.

Somebody else reported some ghost WU's in the Cafe here.
I have had one as well. As Mod.Sense says, this is nothing to worry about.

---

This is interesting. The researcher who submitted these jobs has not received any results yet, but the status of the tasks show that they have been sent to clients.

There may be an issue here that we have to look into. These tasks have a large number of input files relative to other jobs.

Please post if you are having the same issues.

Hi,

I have been getting major download problems.

This from one box

19/05/2008 23:11:09|rosetta@home|Sending scheduler request: Requested by user. Requesting 352238 seconds of work, reporting 0 completed tasks
19/05/2008 23:11:14|rosetta@home|[error] Can't parse workunit in scheduler reply: unexpected XML tag or syntax
19/05/2008 23:11:14|rosetta@home|[error] No close tag in scheduler reply
19/05/2008 23:12:14|rosetta@home|Sending scheduler request: Requested by user. Requesting 352554 seconds of work, reporting 0 completed tasks
19/05/2008 23:12:19|rosetta@home|Scheduler request succeeded: got 0 new tasks
19/05/2008 23:12:19|rosetta@home|Message from server: Not sending work - last request too recent: 66 sec
19/05/2008 23:16:24|rosetta@home|Sending scheduler request: To fetch work. Requesting 353752 seconds of work, reporting 0 completed tasks
19/05/2008 23:16:34|rosetta@home|[error] Can't parse workunit in scheduler reply: unexpected XML tag or syntax
19/05/2008 23:16:34|rosetta@home|[error] No close tag in scheduler reply
19/05/2008 23:18:19|rosetta@home|Sending scheduler request: To fetch work. Requesting 354323 seconds of work, reporting 0 completed tasks
19/05/2008 23:18:39|rosetta@home|Scheduler request succeeded: got 0 new tasks
19/05/2008 23:18:39|rosetta@home|Message from server: Not sending work - last request too recent: 130 sec
19/05/2008 23:22:44|rosetta@home|Sending scheduler request: To fetch work. Requesting 355581 seconds of work, reporting 0 completed tasks
19/05/2008 23:22:54|rosetta@home|[error] Can't parse workunit in scheduler reply: unexpected XML tag or syntax
19/05/2008 23:22:54|rosetta@home|[error] No close tag in scheduler reply

Been going on for the last 5 hours that i know of.

Cheers

---

Just noticed this in my message log:

Mon May 19 19:20:32 2008|rosetta@home|Sending scheduler request: To fetch work. Requesting 16 seconds of work, reporting 0 completed tasks
Mon May 19 19:20:38 2008|rosetta@home|[error] Can't parse workunit in scheduler reply: unexpected XML tag or syntax
Mon May 19 19:20:38 2008|rosetta@home|[error] No close tag in scheduler reply

It appears to be this wu, BAK1b0o_loop_model_biased_it18_3335_466 ,which shows up on my tasks page but never made it to my computer - a ghost.

Snags

I just noticed that some of my machines have ghost WUs from around May 15.
For instance: https://boinc.bakerlab.org/rosetta/results.php?hostid=165747
has 4 WUs from May 15 that are ghosts.

The message log has something like the following for each ghost WU:

Sending scheduler request to https://boinc.bakerlab.org/rosetta_cgi/cgi
Reason: To fetch work
Requesting 979 seconds of new work
Scheduler request to https://boinc.bakerlab.org/rosetta_cgi/cgi succeeded
Can't parse work unit in scheduler reply: -112
No close tag in scheduler reply
Can't parse scheduler reply


This is from a Linux system with Boinc 5.2.13

Yep. DK is on to it. There was a problem with how the work units were created that caused the BOINC client to be unable to get off the ground with them. So, there's nothing to change or do on your end. And your client doesn't realize those work units were assigned to it, due to the XML errors, so there aren't any tasks to abort either.

---

Rosetta Moderator: Mod.Sense

Okay, I think I may have fixed the issue. I had to recompile our scheduling server programs with a larger xml_doc max buffer. A few batches were submitted that had so many input files that the scheduler reply from our server was truncated for these tasks. Please ignore the "ghost" work units. They will eventually get cleared.