Result '(result)' exited with zero status but no 'finished' file
From Unofficial BOINC Wiki
Contents |
[edit] General
Message Type: Status Message
This message indicates that the BOINC Client Software is telling the Participant that they may have a problem on their computer. In those cases where the problem is continual the cure may be to "detach" from the Project or to do a Project "Reset". This will let the BOINC Client Software delete all of the files related to that Project so that, hopefully, the bad file will be eliminated. However, most of the time the best thing to do is to do nothing, the BOINC Client Software will normally recover with no intervention by the Participant.
Before you do a Project "Reset" or "detach" from a Project, check with the Participants that assist on the "Questions & Problems" forums of the affected project!
I would further suggest that you do a complete disk check using a good utility to make sure that there are no structural problems on the disk. Note that you must turn the BOINC Client Software off and make sure there are no other programs running that write to the disk. You may, if you are using Microsoft Windows®, have to re-boot to allow the disk checking program access to the disk drive.
Another cause would be when the BOINC Manager dies but does not stop the Science Application before it quits. When this happens, the Science Application will stop with no error, that is what the "zero status" tells us, but since it is not done processing the Work Unit yet, it did not create a Result Data File. In this case, when the system is restarted the BOINC Client Software will restart the processing of the Work Unit at the last Checkpoint.
- Note:
- This is normally considered a "benign" error and in most cases is nothing to get excited about. However, if every Work Unit is showing this message, well, then it is something to be looked into, otherwise, just check and make sure the Work Unit is completing and you are getting a Valid Result.
- Note:
- One of the causes of this message seems to be the setting of the computer's clock. When the time is adjusted the BOINC Daemon and the Science Application seem to get out of step. This should be fixed in the 4.7x/5.0.x release of the BOINC Client Software.
[edit] Detailed Explanation
This error is only a problem if it is continuous, like if every time the Science Application starts it runs for a minute (initializing) and then exits. What you can do is look in the "slots" folder for the stderr.txt file and see what kind of error messages show up in it.
When the Science Application exits, it was because it "lost contact" with the BOINC Daemon. The BOINC Daemon sends "heartbeat" messages to each of the running Science Application(s) that say "I'm alive", when the Science Application doesn't get any for 30 seconds, it exits with a "no heartbeat" message. Check the Result on the web site, it'll show the message in the <stderr> section.
A few of reasons that this can happen include:
- The BOINC Daemon does a DNS lookup, and "hangs" until the response comes back. If your Internet connection isn't working or you're disconnected, and it "hangs" for 30 seconds or more, the Science Applications will all exit. You'll see DNS lookup type errors in the log (like: "Can't resolve hostname ['(host-name)'] host name not found"), followed by the Result restarting.
- Note:
- These network "hangs" trigging the error do not occur for all Participants under these conditions.
- The BOINC Daemon starts a new Result, it checks to make sure it can communicate with the newly started the Science Application. If it can't, it'll "sleep" for 35 seconds and all of the running the Science Applications will exit with "no heartbeat" messages. This is intentional, and makes sure the just-started Science Application is using the correct "shared memory" segment. You'll see messages about the BOINC Daemon is starting a new Result and suspending (preempting) another; and the "exiting with zero status" message occurs at least 30 seconds after that.
- The BOINC Daemon can't write the Client State File, it'll "sleep" for 60 seconds to give the other application time to move out of the way. A side effect is that the Science Applications all exit and the BOINC Daemon has to restart them. You'll see "Couldn't write state file: '(error-code)'" in the message. If you see these, it usually means other programs are interfering with the BOINC Client Software, like Anti-Virus or file indexing utilities.
- File system activity on the Microsoft Windows® system is enough that requests get queued up faster than Microsoft Windows® can service them. Copying a lot of small files from one directory to another is one cause, anything scanning all the files on a disk is another. The BOINC Client Software can "hang" since it checks the state of its files at regular intervals. You'll notice very slow response time when this happens.
The kind of filesystem problems that would cause the BOINC Daemon to lock up aren't usually checked by the disk check programs. You would find them in the system event log, things like "The driver disabled the write cache on device \Device\Harddisk3\DR48" or "The driver detected a controller error on \Device\Harddisk1\D".
Usually these are due to system performance problems. For whatever reason, the BOINC Daemon didn't execute for a long enough period that the Science Applications thought it wasn't running anymore. There could be many reasons, top of the list are interactions between anti-virus, file indexing, and writing cached data from the disk cache to the hard drives. Believe it or not, the graphics driver can also get involved, one manufacturer uses locks quite extensively, and this has a negative impact on system performance.
[edit] Example Log(s)
[edit] Successful Work Unit Completion
In this example, the Work Unit is taken from its initial start through the entire process to the delivery of the Result Data Files to the Project's Data Server.
Note: This is a compressed log in that there were a significant number of starts and stops. The good news is that this is a pretty good example in that it shows processing through a computer restart. I did leave out a few details where it seemed that the BOINC Client Software could not seem to make up its mind. Those places where I left out other messages are noted with "Log Breaks", not that it really matters, but, full disclosure you know ...
2005-06-02 07:29:23 [SETI@home] Starting result 30dc04aa.27301.24610.329810.55_1 using setiathome
version 4.02
2005-06-02 08:16:06 [SETI@home] Pausing result 30dc04aa.27301.24610.329810.55_1 (left in memory)
2005-06-02 08:16:41 [SETI@home] Resuming result 30dc04aa.27301.24610.329810.55_1 using setiathome
version 4.02
2005-06-02 08:17:02 [SETI@home] Result 30dc04aa.27301.24610.329810.55_1 exited with zero status
but no 'finished' file
2005-06-02 08:17:02 [SETI@home] If this happens repeatedly you may need to reset the project.
2005-06-02 09:38:49 [ ] Starting BOINC client version 4.43 for powerpc-apple-darwin
2005-06-02 09:38:50 [SETI@home] Deferring computation for result 30dc04aa.27301.24610.329810.55_1
2005-06-02 10:38:50 [SETI@home] Restarting result 30dc04aa.27301.24610.329810.55_1 using setiathome
version 4.02
2005-06-02 14:54:50 [SETI@home] Computation for result 30dc04aa.27301.24610.329810.55_1 finished
2005-06-02 14:54:51 [SETI@home] Started upload of 30dc04aa.27301.24610.329810.55_1_0
2005-06-02 14:54:53 [SETI@home] Finished upload of 30dc04aa.27301.24610.329810.55_1_0
2005-06-02 14:54:53 [SETI@home] Throughput 369389 bytes/sec
Line-By-Line Explanation
- This message lists the identification data on the Result, Science Application, and Version of the Science Application that are going to be put to work.
- This message tells us that the Work Scheduler has changed processing to another Result or to run the Benchmark, etc. and therefore the processing on this Work Unit/Result is going to be interrupted or paused. The current work will be retained within Virtual Memory for later resumption. This makes for a slightly faster resumption.
- Result '(result)' exited with zero status but no 'finished' file
- If this happens repeatedly you may need to reset the project.
-
Result 30dc04aa.27301.24610.329810.55
- These two lines indicate that the process halted for an unknown reason and the BOINC Client Software was expecting that the Work Unit/Result should have been finished.
- Note: There are some problems with this message and investigation is still under way to identify the causes of these messages. As you will note if you follow this log to the end that this Result was completed (Claimed Credit: 15.46 and Granted Credit: 21.89).
- This message is showing that I had restarted the BOINC Client Software and the program started up.
- This message only appears during BOINC startup. This message indicates that the BOINC Client Software decided based on the current CPU Scheduler mode and the relevant Deadlines and Short Term Debts that this Work Unit/Result would not be restarted at this time as some other Work Unit is a better candidate.
- And now it has restarted processing. There is little difference between "Resuming" and "Restarting", "Resuming" occurs if the Result's information is still resident in memory and the "Restarting" indicates a reload of the in-progress information.
- At last! Done! Now lets upload this bad puppy!
- The BOINC Daemon initiated an upload action with the SETI@Home Project.
- The BOINC Daemon completed the data transfer to the Project's Data Server.
- The measured speed of transfer, in this case was an impressive 369,389 bytes per second. This is a fairly high rate of data transfer and is possible as I have a Cable Modem and an "always-on" Internet connection.
RSS Feeds
