Seti WU not received
5 Jun 2007 3:46:41 UTC
Topic 192814
(moderation:
I recently noticed that Einstein is working continuously and not giving up time to Seti. I probably made things worse by trying different things. However, now I have only a single Einstein work unit processing. Have sent requests to Seti for work and from the messages it appear to be sending units but it does not show up on the tasks tab. Tried to use update on the Seti project but no work received. Any help?
Seti WU not received
)
I suspect that either the Einstein task you have is close to deadline or you have done more work for SETI than Einstein recently. In either case usually the best thing to do is leave BOINC alone and let it straighten itself out.
If neither of these ideas seems applicable post back and we will look into other possiblities.
BOINC WIKI
BOINCing since 2002/12/8
As of 4:20 PM, June 6th,
)
As of 4:20 PM, June 6th, still no SETI work received. My messages are:
6/6/2007 12:21:30 PM||Starting BOINC client version 5.8.16 for windows_intelx86
6/6/2007 12:21:30 PM||log flags: task, file_xfer, sched_ops
6/6/2007 12:21:30 PM||Libraries: libcurl/7.16.0 OpenSSL/0.9.8a zlib/1.2.3
6/6/2007 12:21:30 PM||Data directory: C:\\Seti BOINC
6/6/2007 12:21:30 PM||Processor: 1 GenuineIntel Intel(R) Pentium(R) 4 CPU 1700MHz [x86 Family 15 Model 0 Stepping 10] [fpu tsc sse sse2 mmx]
6/6/2007 12:21:30 PM||Memory: 511.07 MB physical, 1.22 GB virtual
6/6/2007 12:21:30 PM||Disk: 74.50 GB total, 29.67 GB free
6/6/2007 12:21:31 PM|Einstein@Home|URL: http://einstein.phys.uwm.edu/; Computer ID: 840931; location: home; project prefs: default
6/6/2007 12:21:31 PM|SETI@home|URL: http://setiathome.berkeley.edu/; Computer ID: 3032913; location: home; project prefs: home
6/6/2007 12:21:31 PM||General prefs: from SETI@home (last modified 2007-06-03 17:25:20)
6/6/2007 12:21:31 PM||Host location: home
6/6/2007 12:21:31 PM||General prefs: using separate prefs for home
6/6/2007 12:21:31 PM||file projects/einstein.phys.uwm.edu/einstein_4.79_windows_intelx86.exe not found
6/6/2007 12:21:31 PM||file projects/einstein.phys.uwm.edu/einstein_4.79_windows_intelx86.pdb not found
6/6/2007 12:21:31 PM||file projects/einstein.phys.uwm.edu/einstein_S5R1_4.24_windows_intelx86.exe not found
6/6/2007 12:21:31 PM||file projects/einstein.phys.uwm.edu/einstein_S5R1_4.24_windows_intelx86.pdb not found
6/6/2007 12:21:31 PM||file projects/einstein.phys.uwm.edu/einstein_S5RI_4.24_windows_intelx86.exe not found
6/6/2007 12:21:31 PM||file projects/einstein.phys.uwm.edu/einstein_S5RI_4.24_windows_intelx86.pdb not found
6/6/2007 12:21:36 PM|Einstein@Home|Restarting task h1_0405.30_S5R2__141_S5R2c_2 using einstein_S5R2 version 417
6/6/2007 12:21:37 PM|Einstein@Home|[file_xfer] Started download of file einstein_4.79_windows_intelx86.exe
6/6/2007 12:21:37 PM|Einstein@Home|[file_xfer] Started download of file einstein_4.79_windows_intelx86.pdb
6/6/2007 12:21:46 PM|Einstein@Home|[file_xfer] Finished download of file einstein_4.79_windows_intelx86.exe
6/6/2007 12:21:46 PM|Einstein@Home|[file_xfer] Throughput 272241 bytes/sec
6/6/2007 12:21:46 PM|Einstein@Home|[file_xfer] Started download of file einstein_S5R1_4.24_windows_intelx86.exe
6/6/2007 12:21:48 PM|Einstein@Home|[file_xfer] Finished download of file einstein_4.79_windows_intelx86.pdb
6/6/2007 12:21:48 PM|Einstein@Home|[file_xfer] Throughput 349223 bytes/sec
6/6/2007 12:21:48 PM|Einstein@Home|[file_xfer] Started download of file einstein_S5R1_4.24_windows_intelx86.pdb
6/6/2007 12:21:50 PM|Einstein@Home|[file_xfer] Finished download of file einstein_S5R1_4.24_windows_intelx86.exe
6/6/2007 12:21:50 PM|Einstein@Home|[file_xfer] Throughput 407555 bytes/sec
6/6/2007 12:21:50 PM|Einstein@Home|[file_xfer] Started download of file einstein_S5RI_4.24_windows_intelx86.exe
6/6/2007 12:21:56 PM|Einstein@Home|[file_xfer] Finished download of file einstein_S5R1_4.24_windows_intelx86.pdb
6/6/2007 12:21:56 PM|Einstein@Home|[file_xfer] Throughput 525242 bytes/sec
6/6/2007 12:21:56 PM|Einstein@Home|[file_xfer] Started download of file einstein_S5RI_4.24_windows_intelx86.pdb
6/6/2007 12:21:58 PM|Einstein@Home|[file_xfer] Finished download of file einstein_S5RI_4.24_windows_intelx86.exe
6/6/2007 12:21:58 PM|Einstein@Home|[file_xfer] Throughput 204868 bytes/sec
6/6/2007 12:22:08 PM|Einstein@Home|[file_xfer] Finished download of file einstein_S5RI_4.24_windows_intelx86.pdb
6/6/2007 12:22:08 PM|Einstein@Home|[file_xfer] Throughput 326863 bytes/sec
Any ideas why SETI not providing work or I'm not getting work?
Have you been resetting the
)
Have you been resetting the Einstein project? Seeing it redownload executables is a bit unusual. Resetting Einstein could also be causing your client to ignore SETI.
Please look in the client_state.xml file in your BOINC directory for the tags long_term_debt. There should be one for SETI and one for Einstein. Let us know what is there. They should be +- your connect setting in seconds.
BOINC WIKI
BOINCing since 2002/12/8
This is what is shown for
)
This is what is shown for Seti in long_term_debt:
53938.659126
Under Einstein, it shows:
-53938.659126
It seems odd that both would contain the same values. client_state.xml has a lot of information in it but it mainly relates to Einstein. It is 878 lines long with 28 lines specifically referring to Seti with the rest referring mainly to Einstein and some to just the computer configuration info.
Appreciate any help you can give me.
That's the way BOINC works,
)
That's the way BOINC works, the debt values are normalized so that the mean is zero.
Looking at your LTD, the next result which gets DL'ed should come from SAH. You have your host hidden on SAH, but I'm going to assume you don't have any results for them at this point. So the most likely reason you haven't DL'ed one yet is a low Uptime value (IOW, you don't run the host 24/7) and you just haven't gotten to the next work fetch point set by your CI.
Alinator
Just for reference, work fetch is determined by the LTD. Work will be DL'ed from active projects with work to send which has the greatest relative LTD value.
Not always certain as to your
)
Not always certain as to your meaning due to terminology. I run one computer at home. Seti member since 1999 and Einstein since 2005. With one computer, not much computing power to donate. Previous to this problem, both projects would have work to be done showing in Tasks tab. Now only Einstein shows work to be done. If I say update to Seti, message reflects I'm not reporting results or requesting new work which is exactly what I'm doing.
RE: I recently noticed that
)
Just a thought here that "MIGHT" explain it. I've had the same problem with one of my systems as well and I think the problem came because the einstein wu was going to take almost 2 days. I think boinc figured it was already filled for the day and didn't download. I gave it a day and was thinking about putting einstein on "No New Tasks" until seti got one to work on when I saw it called out to seti for a wu. When it did it got 2 of them.
If it hasn't worked itself out in a day or so I'd set einstein to No New Tasks and leave it like that until after seti got some work and see if that fixes it.
HTH....
RE: That's the way BOINC
)
Alinator's explanation is exactly right. Let me see if I can explain without the jargon. First 50K seconds is between 1 and 2 days. So there does not appear to be a problem, your client just needs to finish the einstein tasks on hand. There is/was probably a deadline issue that is preventing work fetch. The next time the client is ready for more work it should try SETI first.
When BOINC is figuring how much work to get it figures in how often your computer is on and how often you allow processing of BOINC tasks. So a 2 day setting for "connect every x days" may get less work than you expect but it will normally be fairly accurate about when the work is returned.
Trying to force the client to do what you expect by seting no new tasks or suspending tasks/projects is usually not a good idea. In most cases it makes it harder for your client to get to a balanced state and therefore less likely to do what you expect it to do. In fact the thing to do if you don't get SETI work when the einstein work on hand is finished is to check to see if the SETI project is suspended or set to no new tasks. Remember the buttons in the manager say what will happen if you push them so they should say "suspend" or "no new tasks".
BOINC WIKI
BOINCing since 2002/12/8
RE: Not always certain as
)
LOL....
No problemo, sometimes I don't always look at the number of posts someone has made and resort to using the 'alphabet' soup common acronyms.
CI = Your "Connect to Network" interval. This sets the approximate size of the work cache for BOINC. Note, this will change somewhat with the next major release of BOINC.
LTD = "Long Term Debt". This determines which active project will get work on the next work fetch cycle.
STD = "Short Term Debt". This determines which task will run next, as long as there are no other potential scheduling or deadline problems with the other tasks onboard.
As noted before, these are normalized so the mean is zero.
Uptime = This is the percent of time BOINC is running metric from your Computer Summary. Note, this is not necessarily the same as the host's OS reported uptime unless you run BOINC as a service (or some other technique for starting it automatically on some platforms, like Win 9x for example).
Note, the other project monitored performance metrics (listed after after Uptime) are show at the bottom of the Computer Summary page. They include in order:
BAT = "BOINC Allowed Time" (in percent). A measure of how much time the BOINC Core Client is actually allowed to run. This is effected by operational preference restrictions you put on BOINC.
ACE = "Average CPU Efficiency". A measure of how busy your host is doing other non-BOINC related tasks.
RDCF = "Result Duration Correction Factor". A derived metric used to compensate for the inaccuracies inherent in all 'synthetic' benchmarks when used to characterize a hosts performance in 'real world' computational scenarios. It basically compares how your host actually performs on the results to what the project thinks it should be able to do based on the benchmark estimates. This is used to more accurately determine how much work can be carried for a given work cache setting.
The bottom line is that the Core Client and the Project side Work Schedulers use all of this data in determining when, how much, and from where work is to be drawn from on every work fetch cycle. The 'Supreme' guiding factor in making these decisions is that all work onboard should be returned before the deadline for it expires. Since some of the parameters can be somewhat mutally exclusive when looked at short term, BOINC is designed to take as few chances as possible when it comes to taking on new work. This can frequently lead to it's behaviour diverging from what a human would do when presented with the same 'facts', but over the long haul results in it honoring the preferences you have set for it nonetheless.
HTH,
Alinator