Problem reporting tasks


log in

Advanced search

Message boards : Number crunching : Problem reporting tasks

1 · 2 · 3 · 4 . . . 7 · Next
Author Message
Ronald
Send message
Joined: 15 May 99
Posts: 8
Credit: 1,562,838
RAC: 0
United Kingdom
Message 1020128 - Posted: 27 Jul 2010, 15:04:54 UTC

Hi. I seem to be having a problem reporting tasks from one of my machines (currently 30 tasks)

I get the following messages :

scheduler request failed: HTTP internal server error

or

scheduler request failed: Couldn't connect to server
Internet access OK - project servers may be temporarily down

It has no problems uploading tasks

This machine is using an anonymous platform

cheers

dlanor
____________

Profile hiamps
Volunteer tester
Avatar
Send message
Joined: 23 May 99
Posts: 4292
Credit: 72,971,319
RAC: 0
United States
Message 1020129 - Posted: 27 Jul 2010, 15:06:11 UTC - in response to Message 1020128.

Hi. I seem to be having a problem reporting tasks from one of my machines (currently 30 tasks)

I get the following messages :

scheduler request failed: HTTP internal server error

or

scheduler request failed: Couldn't connect to server
Internet access OK - project servers may be temporarily down

It has no problems uploading tasks

This machine is using an anonymous platform

cheers

dlanor

Most of us are the servers are maxed getting ready to shut down. Peole are trying to get last minute work.
____________
Official Abuser of Boinc Buttons...
And no good credit hound!

OfficeDude
Send message
Joined: 7 Jul 99
Posts: 8
Credit: 1,847,065
RAC: 320
Belgium
Message 1020131 - Posted: 27 Jul 2010, 15:09:38 UTC

This message already started to appear in my logs 8 hours ago. Ooh well, the early bird gets....
____________

Ronald
Send message
Joined: 15 May 99
Posts: 8
Credit: 1,562,838
RAC: 0
United Kingdom
Message 1020132 - Posted: 27 Jul 2010, 15:15:33 UTC - in response to Message 1020129.

Most of us are the servers are maxed getting ready to shut down. Peole are trying to get last minute work.


That's a pity as some of it will have gone past the report deadline before i'll be able to get back online.

thanks

dlanor

____________

Profile hiamps
Volunteer tester
Avatar
Send message
Joined: 23 May 99
Posts: 4292
Credit: 72,971,319
RAC: 0
United States
Message 1020136 - Posted: 27 Jul 2010, 15:29:57 UTC - in response to Message 1020132.

Most of us are the servers are maxed getting ready to shut down. Peole are trying to get last minute work.


That's a pity as some of it will have gone past the report deadline before i'll be able to get back online.

thanks

dlanor

You have a bit of time yet, keep hitting update I got a few in....
____________
Official Abuser of Boinc Buttons...
And no good credit hound!

Profile hiamps
Volunteer tester
Avatar
Send message
Joined: 23 May 99
Posts: 4292
Credit: 72,971,319
RAC: 0
United States
Message 1020138 - Posted: 27 Jul 2010, 15:35:10 UTC

Didn't help that the first project backoff for my machine started at 1 hour 12 minutes and went up from there. If boinc couldn't report sometimes it would try and back off for 2 or 3 hours...With the small amount of time we have to report and get work, project backoffs should be less not more!
____________
Official Abuser of Boinc Buttons...
And no good credit hound!

Profile hiamps
Volunteer tester
Avatar
Send message
Joined: 23 May 99
Posts: 4292
Credit: 72,971,319
RAC: 0
United States
Message 1020139 - Posted: 27 Jul 2010, 15:36:09 UTC - in response to Message 1020132.

Most of us are the servers are maxed getting ready to shut down. Peole are trying to get last minute work.


That's a pity as some of it will have gone past the report deadline before i'll be able to get back online.

thanks

dlanor

7/27/2010 8:32:40 AM | SETI@home | update requested by user

7/27/2010 8:32:44 AM | SETI@home | Sending scheduler request: Requested by user.

7/27/2010 8:32:44 AM | SETI@home | Reporting 5 completed tasks, requesting new tasks for CPU and GPU

7/27/2010 8:33:08 AM | SETI@home | Scheduler request completed: got 0 new tasks


____________
Official Abuser of Boinc Buttons...
And no good credit hound!

JohnDKProject donor
Volunteer tester
Avatar
Send message
Joined: 28 May 00
Posts: 862
Credit: 46,461,525
RAC: 71,988
Denmark
Message 1020155 - Posted: 27 Jul 2010, 16:02:03 UTC - in response to Message 1020129.

Most of us are the servers are maxed getting ready to shut down. Peole are trying to get last minute work.

Cricket shows the last hour not being maxed out, still there's often no connect or no tasks available. This has been going on I think most of the time after last outage, I *think* there's some other problem going on.

Profile hiamps
Volunteer tester
Avatar
Send message
Joined: 23 May 99
Posts: 4292
Credit: 72,971,319
RAC: 0
United States
Message 1020159 - Posted: 27 Jul 2010, 16:09:27 UTC
Last modified: 27 Jul 2010, 16:11:05 UTC

The last hour I got quite a bit of work, probably will need a new client as my "update" button is probably getting pretty worn...LOL But with that and the reschedule I might make it.

EDIT:
7/27/2010 9:09:43 AM | SETI@home | update requested by user

7/27/2010 9:09:44 AM | SETI@home | Sending scheduler request: Requested by user.

7/27/2010 9:09:44 AM | SETI@home | Reporting 3 completed tasks, requesting new tasks for CPU and GPU

7/27/2010 9:10:15 AM | SETI@home | Scheduler request completed: got 0 new tasks

7/27/2010 9:10:15 AM | SETI@home | Message from project server: Project has no tasks available

7/27/2010 9:10:30 AM | SETI@home | Sending scheduler request: To fetch work.

7/27/2010 9:10:30 AM | SETI@home | Requesting new tasks for CPU
____________
Official Abuser of Boinc Buttons...
And no good credit hound!

Profile S@NL - XP_Freak
Send message
Joined: 10 Jul 99
Posts: 99
Credit: 4,838,765
RAC: 1,947
Netherlands
Message 1020169 - Posted: 27 Jul 2010, 16:30:24 UTC
Last modified: 27 Jul 2010, 16:33:05 UTC

I have not been able to report work since yesterday 26-06-2010 20:00 UTC
That's now for about 20 hours.
I did not get any new WU's during that time.
Therefore I think I will be out of work in a little more then two days.


27-7-2010 17:47:41 SETI@home Reporting 3 completed tasks, requesting new tasks
27-7-2010 17:48:04 SETI@home Scheduler request failed: HTTP internal server error
____________

Goodbye Seti Classic

Bearcat
Send message
Joined: 10 Sep 99
Posts: 106
Credit: 10,778,506
RAC: 0
United States
Message 1020170 - Posted: 27 Jul 2010, 16:31:45 UTC - in response to Message 1020159.
Last modified: 27 Jul 2010, 16:33:54 UTC

Cricket shows the last hour not being maxed out, still there's often no connect or no tasks available. This has been going on I think most of the time after last outage, I *think* there's some other problem going on.


That problem has been explained a few times before, but I cannot find the right thread right now. It has something to do with the scheduler or the feeder only having 100 units every 6 seconds, and if those are used/taken/send to someone else, you'll get the msg no work available althought there might be thousands of units shown on the server status page or the cricket graph not being maxed out. In other words, keep trying , it'll work eventually :)
____________

Profile hiamps
Volunteer tester
Avatar
Send message
Joined: 23 May 99
Posts: 4292
Credit: 72,971,319
RAC: 0
United States
Message 1020180 - Posted: 27 Jul 2010, 21:53:14 UTC - in response to Message 1020169.

I have not been able to report work since yesterday 26-06-2010 20:00 UTC
That's now for about 20 hours.
I did not get any new WU's during that time.
Therefore I think I will be out of work in a little more then two days.


27-7-2010 17:47:41 SETI@home Reporting 3 completed tasks, requesting new tasks
27-7-2010 17:48:04 SETI@home Scheduler request failed: HTTP internal server error

Yep, they are making you work for them now a days....
____________
Official Abuser of Boinc Buttons...
And no good credit hound!

JohnDKProject donor
Volunteer tester
Avatar
Send message
Joined: 28 May 00
Posts: 862
Credit: 46,461,525
RAC: 71,988
Denmark
Message 1020368 - Posted: 28 Jul 2010, 13:32:09 UTC - in response to Message 1020170.

That problem has been explained a few times before, but I cannot find the right thread right now. It has something to do with the scheduler or the feeder only having 100 units every 6 seconds, and if those are used/taken/send to someone else, you'll get the msg no work available althought there might be thousands of units shown on the server status page or the cricket graph not being maxed out. In other words, keep trying , it'll work eventually :)

Yes I know about that, I just think it has been happening much more lately, as if there's some other problem going on.

Josef W. SegurProject donor
Volunteer developer
Volunteer tester
Send message
Joined: 30 Oct 99
Posts: 4298
Credit: 1,067,168
RAC: 1,010
United States
Message 1020424 - Posted: 28 Jul 2010, 18:43:44 UTC - in response to Message 1020155.

Most of us are the servers are maxed getting ready to shut down. Peole are trying to get last minute work.

Cricket shows the last hour not being maxed out, still there's often no connect or no tasks available. This has been going on I think most of the time after last outage, I *think* there's some other problem going on.

The "no tasks available" part is fairly easy. There are long stretches of the resultID sequence which have no AP tasks, so the Feeder only fills the 96 (my guess) slots for MB tasks. Those 96 represent about 296 Mbits to be downloaded, and if the Feeder is using the default 5 second sleep periods then the average download rate cannot be more than about 59.2 Mbits/second. Meanwhile, those 96 get assigned to hosts quickly, and during the remainder of the 5 second sleep period other hosts get "no tasks available". Note that parts of the resultID sequence which do have 3 AP tasks for each Feeder run add about 206.5 Mbits to the downloads each time, giving somewhat more than 100 Mbits/sec assigned download rate and leading to those bursts where the line is saturated.

The "no connect" is much more difficult to diagnose. Essentially it just means that the handshaking required to establish a connection is failing someplace. Given about 260000 active hosts and estimating each trying to get work 4 times a day, that's 12 scheduler requests a second and shouldn't be difficult to handle. But when a host does get a successful work request in which gets less work than the host wants, the request backoff is cleared and the host does additional requests at one minute intervals. In effect more successful requests may greatly increase the overall rate of requests. I don't know whether that contributes to connect failures, but it seems it might.
Joe

Ingleside
Volunteer developer
Send message
Joined: 4 Feb 03
Posts: 1546
Credit: 4,332,353
RAC: 1,528
Norway
Message 1020489 - Posted: 29 Jul 2010, 0:47:46 UTC - in response to Message 1020424.

But when a host does get a successful work request in which gets less work than the host wants, the request backoff is cleared and the host does additional requests at one minute intervals.

If host gets some work on a work-request, there's only the project-specified backoff, this is atleast 6 seconds (if project haven't specified anything). For SETI@home the deferral is 11 seconds, so a host can max manage making 5 work-requests per minute. In practice every step in a scheduler-request will take a little time, so would guess 3 work-requests per minute is the practical max, if all of them is successful that is.



____________
"I make so many mistakes. But then just think of all the mistakes I don't make, although I might."

Profile Searcher
Avatar
Send message
Joined: 26 Jun 99
Posts: 139
Credit: 5,388,579
RAC: 1,032
United States
Message 1020912 - Posted: 30 Jul 2010, 19:42:02 UTC - in response to Message 1020489.
Last modified: 30 Jul 2010, 19:42:16 UTC

Looks like I have been getting the same error as OP.
I have 57 ready to report.
:(

S@NL - John van GorselProject donor
Volunteer tester
Avatar
Send message
Joined: 5 Jul 99
Posts: 190
Credit: 137,587,781
RAC: 7,625
Netherlands
Message 1021131 - Posted: 31 Jul 2010, 7:53:43 UTC

I still get the HTTP internal server error or Couldn't connect to server on two of my pc's. Both pc's have not been able to report any tasks since Monday (30 hours before the server outage).
Uploads worked fine on these pc's and 4 other pc's on the same physical network have no problem reporting.

The things I tried so far:
- Be patient and wait for network traffic jam to clear
- Upgrade to Boinc 6.10.58
- Restart the pc
- ipconfig /flushdns

None of it helped and I really would like to report those 523 work units..

Any suggestions on how to diagnose this?
____________


Seti@Netherlands website

Profile S@NL - XP_Freak
Send message
Joined: 10 Jul 99
Posts: 99
Credit: 4,838,765
RAC: 1,947
Netherlands
Message 1021132 - Posted: 31 Jul 2010, 8:08:45 UTC - in response to Message 1021131.
Last modified: 31 Jul 2010, 8:09:39 UTC

I still get the HTTP internal server error or Couldn't connect to server on two of my pc's. Both pc's have not been able to report any tasks since Monday (30 hours before the server outage).
Uploads worked fine on these pc's and 4 other pc's on the same physical network have no problem reporting.

The things I tried so far:
- Be patient and wait for network traffic jam to clear
- Upgrade to Boinc 6.10.58
- Restart the pc
- ipconfig /flushdns

None of it helped and I really would like to report those 523 work units..

Any suggestions on how to diagnose this?

No, I have no suggestions.

I have got the same problem and tried the same things you did, except IPconfig.
____________

Goodbye Seti Classic

Profile Tim Norton
Volunteer tester
Avatar
Send message
Joined: 2 Jun 99
Posts: 835
Credit: 33,540,164
RAC: 0
United Kingdom
Message 1021135 - Posted: 31 Jul 2010, 8:22:27 UTC
Last modified: 31 Jul 2010, 8:28:57 UTC

Hi John

looking at you PC are we talking about linux or windows? - i guess windows

few ideas or thinks to check - you may have already done this :)

i assume you do have network activity enabled and no odd settings in your preferences in BM etc

do you have any other boinc projects on the affected machines and do the report/connect ok? and you can connect to the internet from the affected machines ok? affected machines can see and be seen by others on your network etc

as you have a reasonably large number to report - if you suspend most of them an leave a few to report and see if that makes any difference - unlikely but...

one thing you might try is power cycle you router as it may be a glitch but as your other pc's are ok - unlikley but...


see how you go with that :)


[edit] sorry the suspend idea does not work - thought i had done this successfully myself - lol [/edit]
____________
Tim

Profile S@NL - XP_Freak
Send message
Joined: 10 Jul 99
Posts: 99
Credit: 4,838,765
RAC: 1,947
Netherlands
Message 1021143 - Posted: 31 Jul 2010, 8:48:09 UTC - in response to Message 1021135.

Tim,

The problem of not being able to report started on 26 of July.
That is true for Ronald, OfficeDude, John van Gorsel and me.
Same problem is also reported by others on our forum of SETI@Netherlands.

So I don't think the problem is in our PC's.

This is something that Berkeley has to solve.
____________

Goodbye Seti Classic

1 · 2 · 3 · 4 . . . 7 · Next

Message boards : Number crunching : Problem reporting tasks

Copyright © 2014 University of California