129.74.85.49 - Server does not have record of this unit. Wil

[lanbrown]

The machine that downloaded this unit is the same one that completed it and tried to send the results back. The server is rejecting it for the following reason:
Server does not have record of this unit. Will try again later.

Code: Select all

[20:56:52] - Couldn't send HTTP request to server
[20:56:52] + Could not connect to Work Server (results)
[20:56:52]     (129.74.85.48:8080)
[20:56:52] + Retrying using alternative port
[20:56:52] Connecting to http://129.74.85.48:80/
[20:57:13] - Couldn't send HTTP request to server
[20:57:13] + Could not connect to Work Server (results)
[20:57:13]     (129.74.85.48:80)
[20:57:13] - Error: Could not transmit unit 05 (completed February 26) to work s
erver.
[20:57:13] - 7 failed uploads of this unit.


[20:57:13] + Attempting to send results [February 26 20:57:13 UTC]
[20:57:13] - Reading file work/wuresults_05.dat from core
[20:57:13]   (Read 1722040 bytes from disk)
[20:57:13] Connecting to http://129.74.85.49:8080/
[20:57:30] Posted data.
[20:57:30] Initial: 0000; - Uploaded at ~98 kB/s
[20:57:30] - Averaged speed for that direction ~229 kB/s
[20:57:30] - Server does not have record of this unit. Will try again later.
[20:57:30]   Could not transmit unit 05 to Collection server; keeping in queue.

[7im]

http://fahwiki.net/index.php/Common_Err ... _this_unit

[bruce]

[20:56:52] + Could not connect to Work Server (results)
[20:56:52] (129.74.85.48:8080)

That server has been set to standby. Maybe it's being updated to the new server code and if that's true, it'll probably be back on-line tonight -- but that's just a guess. I'm not sure why.

[lanbrown]

http://fahwiki.net/index.php/Common_Err ... _this_unit

USELESS!!!!!!! I was reporting an issue. If no one ever reports issues, they can't be fixed.

[7im]

Thank you for the issue report.

However, the problem with a collection server not having a record of the work unit is not typically considered an error message, as the fah wiki indicates. The wiki explains why you received that status message (not error message) and that if are simply more patient, the server will come back online, and the work unit will upload automatically. So, IMO, not useless.

And according to Bruce's best guess, the server is in standby status instead of offline status, that is a purposeful status (IMO), not an accidental or failure based status. And as Bruce has guessed, that would seem to indicate the standby is temporary and shortlived. And if not shortlived, it will surely be fixed by Monday morning. And since this is a classic work unit, that work unit isn't going to expire for several months... plenty of time for the issue with the work server to be resolved, and/or the records on the collection server to be updated so your WU can upload, and more importantly, receive credit.

[ONE-OF-THREE]

viewtopic.php?f=18&t=13434&start=120#p131402

Was that "Wiki" article written by someone from by the Pande Group, as it seems odd that you advise that a person should be "simply more patient", while Vijay Pande has "escalated this bug to the highest level" and he, Joe and likely others have worked countless days, nights and weekends trying to fix it.

[7im]

WIKI articles are typically written by seasoned fah contributors with experience on that topic, and a few rare articles are from PG members.

And at first glance, this does appear to be a similar issue. But as Vijay wrote in that same post you referenced...

"Note that, as far as we can tell, this is only an issue for people with multiple GPUs in the same box"

I interpret that to mean the specific issue they put at the top of the list is only a problem with GPU work units. That referenced thread also listed specific GPU work servers, and not the same one as lanbrown. And while lanbrown did not list a specific PRCG number, he did list a work server IP, 129.74.85.48. This is listed as a "classic" server, i.e. CPU work units, unless I am mistaken.

Alternately, if I am completely mistaken, your post showing Vijay attacking this at the highest level results in the same advice to lanbrown. Since the problem is already being attacked at the highest level, all we can do now is be more patient while they fix it.


One thing we might want to ask PG is why no Collection Server is listed for Work Server 129.74.85.48 on the Server Status page? And then why is 129.74.85.49 acting as a Collection Server when it is clearly listed as a Work Server on the Server Status page. New type of virtualized WS/CS combo?

[shdbcamping]

7im,

However, the problem with a collection server not having a record of the work unit is not typically considered an error message, as the fah wiki indicates. The wiki explains why you received that status message (not error message) and that if are simply more patient, the server will come back online, and the work unit will upload automatically. So, IMO, not useless.

I have WU's in my queues Many times that get deleted as the Time has expired for turn-in. All have received this message at finish. I do not know how many Did transmit cause there is no way I can waste that much of my time following a "bug". I could only recommend that you go a little bit easier on those that are contributing to the project but voicing their issues. No one has all the answers and the wiki's are simply generalizations. Good for explanation but offer ZERO solution. If you are so annoyed with folders posting legitimate concerns, please quit replying.

@lanbrown,

As I stated above this problem is far from gone and I am not totally On-board with the trite downplay of the issue in the forums either. If VJay says he's on it, he is. It's all good and most problems are addressed in order of the "Whining" of us, the folding membership. If we don't voice our concerns and weather the scepticism of the forum, the problems will remain.

Sean

[lanbrown]

Thank you for the issue report.

However, the problem with a collection server not having a record of the work unit is not typically considered an error message, as the fah wiki indicates. The wiki explains why you received that status message (not error message) and that if are simply more patient, the server will come back online, and the work unit will upload automatically. So, IMO, not useless.

And according to Bruce's best guess, the server is in standby status instead of offline status, that is a purposeful status (IMO), not an accidental or failure based status. And as Bruce has guessed, that would seem to indicate the standby is temporary and shortlived. And if not shortlived, it will surely be fixed by Monday morning. And since this is a classic work unit, that work unit isn't going to expire for several months... plenty of time for the issue with the work server to be resolved, and/or the records on the collection server to be updated so your WU can upload, and more importantly, receive credit.

Several months? Why not use the server status page? You will see that the longest final deadline is 34.48 days and the shortest being 13.79.

This same server had this same issue not long ago.

I didn't say the article was useless.

[whynot]

I didn't say the article was useless, it was that YOUR post was USELESS!!!!!!

Then, please, the next time elaborate to be more verbose in your statements. I consolidate with 7im in understanding of your post (this one).

EDIT: markup fixes

[ONE-OF-THREE]

lanbrown ...

Now if you would be so kind as to provide a little more information, perhaps someone can help figure out what exactly is the cause of this error message.

So please post the part of the "FAHLog" that shows exactly what type of Work Unit this is, which would for example look something like this "Project: 5799 (Run 1, Clone 34, Gen 1)" and also what type of Folding@Home client you are using to complete these WUs, such a CPU, GPU, or some of the other high-performance clients.

Also it would be extremely helpful to post your F@H name, Team and number, so that someone can hopefully look up your WU to see whether or not it was successfully uploaded or not and if it was, by whom.

Plus since there is so such thing as too much information, perhaps even letting us know your computer specifications might prove to be the last piece of the puzzle in helping to get the answer that you and perhaps others are hoping for.

[noorman]

.

Hear, hear !

.

[bruce]

I have recently experienced the same issue. I processed a WU from Project 10006 and when it was finished, the servers had no record of it. I suspect that this is a remnant of one of the same issues that the FAH servers have been having over the past few weeks and I'm confident that the Pande Group will fix it soon, if they have not fixed it already.

[noorman]

.

On request, I posted a copy of part of a log with such a message; I got a response from V.P. a bit later, that the cause of it was now most probably found and a temp fix had been applied.
The situation was still monitored to check if it worked ...

EDIT: ( I might have been lucky in that I only found 1 such event in several days of Folding on 1 single GPU system )

.

[bruce]

I saw that same "temp fix" comment. I didn't compare the time of my download or my uploads against whatever time that temp fix might have been applied . . . and I don't know which time on my system mattered . . . but we do know that they're aware of the bug and have been working on it.

Topic closed.