I just leave BOINC running, with hardly a look at what's going on. It's just not that interesting to command my time over other things going on...
But this morning I had a look and found that there are eight "6-hour" SETI work units saying "Uploading", and a 255-hour Astropulse that's "Ready to report" - but its deadline was 2 days ago!
I don't know when the uploading will happen (I assume there's something wrong with the servers) but does that mean that the AP unit is going to be ignored and I won't get credit for it? Since I run SETI 50:50 with Einstein (started doing that at a previous long-running SETI outage) that represents over three weeks' worth of electricity and CPU cycles for nothing - this seems ludicrous. Nothing I have done caused this - this machine has been on 24/7 during that time and I don't do any heavy processing, so it's BOINC/SETI itself that has caused this to happen. I'm not *that* upset at losing the credit, but I'm furious that the effort expended won't contribute to the project.

And what's more, could it explain the downturn in progress? It may not be people leaving, but lots of missed deadlines meaning that the work was done, but not acknowledged?

Cheers,
Howard

---

I just leave BOINC running, with hardly a look at what's going on. It's just not that interesting to command my time over other things going on...
But this morning I had a look and found that there are eight "6-hour" SETI work units saying "Uploading", and a 255-hour Astropulse that's "Ready to report" - but its deadline was 2 days ago!
I don't know when the uploading will happen (I assume there's something wrong with the servers) but does that mean that the AP unit is going to be ignored and I won't get credit for it? Since I run SETI 50:50 with Einstein (started doing that at a previous long-running SETI outage) that represents over three weeks' worth of electricity and CPU cycles for nothing - this seems ludicrous. Nothing I have done caused this - this machine has been on 24/7 during that time and I don't do any heavy processing, so it's BOINC/SETI itself that has caused this to happen. I'm not *that* upset at losing the credit, but I'm furious that the effort expended won't contribute to the project.

And what's more, could it explain the downturn in progress? It may not be people leaving, but lots of missed deadlines meaning that the work was done, but not acknowledged?

Cheers,
Howard

This is the wrong forum for this, but I will go over it anyway. Looking at your account, I see an AMD Athlon(tm) XP 2200+ running Microsoft Windows 2000 Professional Edition. I don't see anything that is overdue, nor do I see any Astropulse WU in your list.

Do you have another computer that may be attached to a different account?
I think this issue would be better be addressed in the Q&A section.

(The moderators moved this to Q&A)

Yes, that's the machine.

I don't know where the AstroPulse unit has gone - it was running for a number of weeks, and as I said, last weekend it was showing "Ready to report" - it's not showing on my system any more, and the "Your Results" page doesn't show it at all (is there a way to look back beyond what that shows?). The eight "6 hour" units are there, all sent at the same time earlier today, when presumably whatever was holding that up was resolved.

The Statistics tab here was showing a straight-line descent from the 21st of April to now, until I did a manual communication just now and it's now upwards, obviously including the 8 units sent earlier, but as to where the AP unit has gone, I have no idea - it's just disappeared!

Cheers,
Howard

---

Since the AP task had been over the deadline for some time, probably the other wingmen had reported earlier and the work unit was validated and, after 24 hours, deleted from the database, even before your result was reported.

So, yes, your AP result has vanished - into the wastebin.

Gruß,
Gundolf

[edit] Did you find out why the upload was pending - "Network activity suspended" perhaps?

Are there any messages regarding the upload of the AP result in your Messages tab (stdoutdae.txt)? [/edit]

---

Computer sind nicht alles im Leben. (Kleiner Scherz)

SETI@home classic workunits 3,758
SETI@home classic CPU time 66,520 hours

I haven't stopped network activity - in fact I hadn't touched BOINC for some time at all!
As for it being over the deadline "for some time" - when I looked it was less than 2 days over.

But looking in the log file, I find:

2009-04-21 07:31:18 [SETI@home] Computation for task ap_20dc08ag_B6_P0_00135_20090124_23731.wu_4 finished
...
2009-04-21 07:31:20 [SETI@home] [file_xfer] Started upload of file ap_20dc08ag_B6_P0_00135_20090124_23731.wu_4_0
2009-04-21 07:31:25 [SETI@home] [file_xfer] Finished upload of file ap_20dc08ag_B6_P0_00135_20090124_23731.wu_4_0
2009-04-21 07:31:25 [SETI@home] [file_xfer] Throughput 22944 bytes/sec

So that suggests it *did* get uploaded, three days before the deadline - but looking at my Stats I didn't get credited, my RAC is only 35, and that seems to be from the ordinary units.

So what happened? Why didn't I get the credit (what is the approximate expected credit from an AP unit)?

Why was it still on my machine 7 days after the upload?

Cheers,

Howard

---

You must have an alternate account that you are crunching with, because this account you are posting with does not show an AP workunit having been downloaded in the first place. The fact that you successfully crunched an AP workunit and uploaded some finished file indicates that you were sent an AP workunit to some account, just not the one you are posting here with.


The upload process works in two stages: the BOINC client immediately uploads a file when the workunit is finished to let the servers know that it was completed. The second part actually uploads the result file and a database transaction occurs to begin the validation process only if your quorum partner has returned their result in as well, otherwise the validation awaits the quorum partner's returned result. Credit does not get granted until a quorum has been met. The second part of the upload is delayed so that uploaded results can be batched together to ease the load on the database server across 200,000 clients. The delay usually only lasts for 24 hours.

If the AP WU is still on your machine, can you please provide a WU ID or result ID so that we can track this down?

---

If I do have another account, it's news to me! The AP unit has gone from here, leaving no trace that I can find except in the message log.

A couple of units currently on the account in question are:
Task ID Work Unit
1216698224 438386257 (this one has finished & is pending credit)
1217047009 438543840 (this one is in progress)
Does this help to confirm that we're talking about the same account?

Looking at the log, the AP unit finished on the 21st and was uploaded. But later there seems to be a comms problem, because patterns like the one below seem to be happening for many days - until the 28th, I believe:

2009-04-27 02:56:23 [SETI@home] Sending scheduler request: Requested by user
2009-04-27 02:56:23 [SETI@home] Reporting 1 tasks
2009-04-27 02:56:25 [---] Project communication failed: attempting access to reference site
2009-04-27 02:56:28 [---] Access to reference site succeeded - project servers may be temporarily down.
2009-04-27 02:56:29 [SETI@home] Scheduler request failed: couldn't connect to server
2009-04-27 02:56:29 [SETI@home] Deferring communication for 36 min 17 sec
2009-04-27 02:56:29 [SETI@home] Reason: scheduler request failed

Now if the comms didn't start working until the 28th, and the AP unit's deadline was the 24th, what would happen to the unit when comms resumed (bearing in mind it was uploaded on the 21st)?

Since the unit disappeared on the 28th, the only identity I have for it is the one in the log:

ap_20dc08ag_B6_P0_00135_20090124_23731.wu_4_0

I could send you the whole log (not sure how) if that would help?

Cheers,
Howard

---

If I do have another account, it's news to me! The AP unit has gone from here, leaving no trace that I can find except in the message log.

A couple of units currently on the account in question are:
Task ID Work Unit
1216698224 438386257 (this one has finished & is pending credit)
1217047009 438543840 (this one is in progress)
Does this help to confirm that we're talking about the same account?

Looking at the log, the AP unit finished on the 21st and was uploaded. But later there seems to be a comms problem, because patterns like the one below seem to be happening for many days - until the 28th, I believe:

2009-04-27 02:56:23 [SETI@home] Sending scheduler request: Requested by user
2009-04-27 02:56:23 [SETI@home] Reporting 1 tasks
2009-04-27 02:56:25 [---] Project communication failed: attempting access to reference site
2009-04-27 02:56:28 [---] Access to reference site succeeded - project servers may be temporarily down.
2009-04-27 02:56:29 [SETI@home] Scheduler request failed: couldn't connect to server
2009-04-27 02:56:29 [SETI@home] Deferring communication for 36 min 17 sec
2009-04-27 02:56:29 [SETI@home] Reason: scheduler request failed

Now if the comms didn't start working until the 28th, and the AP unit's deadline was the 24th, what would happen to the unit when comms resumed (bearing in mind it was uploaded on the 21st)?

Since the unit disappeared on the 28th, the only identity I have for it is the one in the log:

ap_20dc08ag_B6_P0_00135_20090124_23731.wu_4_0

I could send you the whole log (not sure how) if that would help?

Cheers,
Howard

More to the point is to match the tasks on the host to the ones on the server that are in progress.

The list aht the server has at the moment is:

07fe09ac.26982.2935.4.8.220_1
07fe09aa.20372.24203.9.8.62_0
01fe09aa.28020.7434.16.8.47_0
01fe09ad.28021.13978.9.8.181_1
07fe09aa.2172.11933.3.8.87_1

If you do not have a subset of these on your computer, you have two different accounts.

You can also check the authenticator in account_setiathome.berkeley.edu.xml against the authenticator on the account you are posting with.

---

BOINC WIKI

I took the WU ID of the finished & pending workunit you gave and cross referenced it with the two accounts it was sent to. Of the two, you are the one awaiting while the other is still in progress, and it is definitely the account you are using here to post with; User ID 57027. So as far as I can tell there is not a second account at play here.


The only two possible scenarios that I can think of to explain this situation is that either the uploaded workunit did not validate and was removed from the system before we had a chance to investigate the problem (all workunits are removed after 24 hours of meeting validation, so it would have had to validate against a third person).

Its either that or this is the first case of a "reverse ghost workunit" where the server sends you an AP and has no record of it (whereas the typical ghost workunit is a task that the server thinks it sends to a host but the host never received the file to work on).

Unfortunately I have no way of tracking down workunits by filename, and once they're removed from the system, as far as I can tell there's no way to track them down at all. As you can see, at this point its all guesswork because we have no way of knowing what happened after it gets removed from the system.

---

OK, well thanks for looking at it. It seems we're at a dead end, and probably will never know what happened.

I'd still like to know why it didn't get sent during the three days between being finished and running out of time - those comms failures were nothing to do with the unit itself, obviously. Was there a major outage that I missed?

Cheers,

Howard

---

The SETI servers, particularly the database server and the forum server (don't know if they're the same physical unit) have had their share of crashes lately. More than likely these were the cause of the comms failures.

---

Well, something is going on, my results page looks like it has been set back about 24 hours. This occured in the last 2 hours. It only shows the last seti enhanced wu's I completed and uploaded yesterday, in green, whereas 2 hours ago it was up to date showing completed. What about the 3 ap wu's I've been crunching for the last 15 hours or so? They don't even show up at all. I would assume it would be pointless to finish them now.

All I know is this is happening regularly usually right after the weekend server maintenance. Anything that manages to upload during maintenance shows up for awhile but tends to disappear the next day, like the server resets itself to the previous day for some reason? Don't know. Hope somebody looks into this, its becoming a little aggravating.

Burn

Hmm, here's an addendum to the above post. I just noticed all my results just came back as they should be, plus 2 of the 3 ap wu's I'm working on now. Maybe that last one will magically return too shortly. Gotta love it!

burn

LOL!...Yep, it did...thanks to whoever fixed it for me. I appreciate it!

burn

Well, something is going on, my results page looks like it has been set back about 24 hours. This occured in the last 2 hours. It only shows the last seti enhanced wu's I completed and uploaded yesterday, in green, whereas 2 hours ago it was up to date showing completed. What about the 3 ap wu's I've been crunching for the last 15 hours or so? They don't even show up at all. I would assume it would be pointless to finish them now.

All I know is this is happening regularly usually right after the weekend server maintenance. Anything that manages to upload during maintenance shows up for awhile but tends to disappear the next day, like the server resets itself to the previous day for some reason? Don't know. Hope somebody looks into this, its becoming a little aggravating.

Burn

I have the same problem and now I'm working on 8 WU that are not listed in the current tasks. Also today I completed about 20 WU and none of them are reported in my task activity. What's going on?

Trucido

Well, something is going on, my results page looks like it has been set back about 24 hours. This occured in the last 2 hours. It only shows the last seti enhanced wu's I completed and uploaded yesterday, in green, whereas 2 hours ago it was up to date showing completed. What about the 3 ap wu's I've been crunching for the last 15 hours or so? They don't even show up at all. I would assume it would be pointless to finish them now.

All I know is this is happening regularly usually right after the weekend server maintenance. Anything that manages to upload during maintenance shows up for awhile but tends to disappear the next day, like the server resets itself to the previous day for some reason? Don't know. Hope somebody looks into this, its becoming a little aggravating.

Burn

I have the same problem and now I'm working on 8 WU that are not listed in the current tasks. Also today I completed about 20 WU and none of them are reported in my task activity. What's going on?

Trucido

Nothing to worry about. See Technical News Forum for explanation.

Yeah,I saw that after the fact, but I still ave 20 wu's from 13 Apr that still show "in progress", that I know were completed and uploaded. So my point is, it's an ongoing problem that really needs a solution if it's going to occur so often. That's a lot of time and work lost. I'm not complaining, just bringing up a point that should be brought to TPTB's attention.

Burn

Yeah,I saw that after the fact, but I still ave 20 wu's from 13 Apr that still show "in progress", that I know were completed and uploaded. So my point is, it's an ongoing problem that really needs a solution if it's going to occur so often. That's a lot of time and work lost. I'm not complaining, just bringing up a point that should be brought to TPTB's attention.

Burn

THe replica DB is behind, and that is where those pages are comming from.

---

BOINC WIKI