I get no WU more

niemz
niemz
Joined: 31 Mar 06
Posts: 1
Credit: 1330
RAC: 0
Topic 191346

Hi all,

After sending the last report on 1.6.06 no more work is done. I checked the scheduler logs and found that:

2006-06-09 05:39:01.8803 [PID=12760] [debug ] REQUEST_METHOD=POST CONTENT_TYPE=application/x-www-form-urlencoded HTTP_ACCEPT=*/* HTTP_USER_AGENT=BOINC client (windows_intelx86 5.4.9)
2006-06-09 05:39:01.8803 [PID=12760] [debug ] CONTENT_LENGTH=2233
2006-06-09 05:39:02.0018 [PID=12760] [normal ] Handling request: host 0, platform windows_intelx86, version 5.4.9, RSF 0.500000
2006-06-09 05:39:02.0019 [PID=12760] [normal ] OS version Microsoft Windows XP Professional Edition, Service Pack 2, (05.01.2600.00)
2006-06-09 05:39:02.0201 [PID=12760] [CRITICAL] [HOST#586023] User has another host with same CPID.
2006-06-09 05:39:02.0213 [PID=12760] [normal ] Processing request [HOST#586023] [RPC#0] core client version 5.4.9
2006-06-09 05:39:02.1429 [PID=12760] [normal ] sending delay request 60.000000

"User has another host with same CPID" - what is that?

Any idea what can i do?

Gary Roberts
Gary Roberts
Moderator
Joined: 9 Feb 05
Posts: 5872
Credit: 117909564667
RAC: 34557841

I get no WU more

Quote:


"User has another host with same CPID" - what is that?

Any idea what can i do?

CPID is the identification number for your machine within the EAH database. In this case it is 586023, also called your Host#. The server seems to think that this number has been doubled up for some reason. I've checked on the website and there is only one machine showing in your list of computers.

As you apparently have no work in progress, you can afford to try a few things without having to worry about losing work. You could try detaching and then reattaching your computer to see if that fixes things. If you get reattached with a different CPID you can "merge" the old ID into the new one and then everything may be back to normal. I've never seen this problem before so I don't really know. Good work for tracking down the relevant entries in the scheduler logs. As a matter of interest, does your BOINC Manager think there is any work actually on your machine?

Try detaching and report back what happens. If necessary, there are other things we can try.

Cheers,

Cheers,
Gary.

PiSA_Fan
PiSA_Fan
Joined: 21 Jun 06
Posts: 4
Credit: 2081296
RAC: 0

RE: "User has another host

Message 36976 in response to message 36975

Quote:


"User has another host with same CPID" - what is that?

Try detaching and report back what happens. If necessary, there are other things we can try.

Cheers,

I've done this but it doesn't helps me. Now I found the following entries in the log

----
2006-06-21 14:00:36.9177 [PID=27847] [debug ]
2006-06-21 14:00:36.9178 [PID=27847] [debug ] REQUEST_METHOD=POST CONTENT_TYPE=application/x-www-form-urlencoded HTTP_ACCEPT=*/* HTTP_USER_AGENT=BOINC client (windows_intelx86 5.4.9)
2006-06-21 14:00:36.9178 [PID=27847] [debug ] CONTENT_LENGTH=3588
2006-06-21 14:00:37.0718 [PID=27847] [normal ] Handling request: host 672428, platform windows_intelx86, version 5.4.9, RSF 0.500000
2006-06-21 14:00:37.0718 [PID=27847] [normal ] OS version Microsoft Windows XP Professional Edition, Service Pack 2, (05.01.2600.00)
2006-06-21 14:00:37.0780 [PID=27847] [debug ] Request [HOST#672428] Database [HOST#672428] Request [RPC#2] Database [RPC#1]
2006-06-21 14:00:37.0787 [PID=27847] [normal ] Processing request [HOST#672428] [RPC#2] core client version 5.4.9
2006-06-21 14:00:37.1505 [PID=27847] [normal ] sending delay request 60.000000
2006-06-21 14:00:36.6734 [PID=27845] [debug ]
----

Why "sending delay request 60.000000" ? Is this a request from server or from client?

Additional I post my local log entries for this request:
----
21.06.2006 16:00:36|Einstein@Home|Sending scheduler request to http://einstein.phys.uwm.edu/EinsteinAtHome_cgi/cgi
21.06.2006 16:00:36|Einstein@Home|Reason: Requested by user
21.06.2006 16:00:36|Einstein@Home|(not requesting new work or reporting completed tasks)
21.06.2006 16:00:41|Einstein@Home|Scheduler request succeeded
----

Here comes allways "not requesting new work or reporting completed tasks". What means this.

By
Falko

Gary Roberts
Gary Roberts
Moderator
Joined: 9 Feb 05
Posts: 5872
Credit: 117909564667
RAC: 34557841

RE: I've done this but it

Message 36977 in response to message 36976

Quote:


I've done this but it doesn't helps me. Now I found the following entries in the log

Why "sending delay request 60.000000" ? Is this a request from server or from client?

This is a routine message sent from the server to the client simply asking the client to wait 60 seconds before making any further requests to the server. Its purpose is to prevent overloading the server. It stops a client from making rapid fire repeated requests, particularly in some error type scenario.

Quote:

Additional I post my local log entries for this request:
----
21.06.2006 16:00:36|Einstein@Home|Sending scheduler request to http://einstein.phys.uwm.edu/EinsteinAtHome_cgi/cgi
21.06.2006 16:00:36|Einstein@Home|Reason: Requested by user
21.06.2006 16:00:36|Einstein@Home|(not requesting new work or reporting completed tasks)
21.06.2006 16:00:41|Einstein@Home|Scheduler request succeeded
----

Here comes allways "not requesting new work or reporting completed tasks". What means this.

It means exactly what it says :).

You initiated the exchange (rather than the BOINC client) by hitting the "Update" button in BOINC Manager. (Requested by user)
The BOINC client is stating that there were no outstanding work requests or reporting requests that it needed to make.

Now the real question is "Why does your client think that it doesn't need more work"?

One possible reason is that you might have work requests disabled in BOINC manager. If you look on the projects tab in BOINC Manager, does it say "Won't get new tasks" or something like that against the E@H project? It's hard to be specific without a lot more information from you.

Here are a couple of observations about situations you should look into. Your first message has certain information under your name at the left which gives your userID as 185089. On the website, your computer has a host# of 586023. Your second message has the same name but a new userID of 200627. Your host# is now 672428. So you have at least two separate accounts in the system. Neither host has any work but the host in your older account has a credit history so at some stage you were crunching successfully. Perhaps you might like to not use your most recent account and try to get the older one working properly.

To allow us to see what is going on, you need to post a lot more of the message log than you did last time. I would like you to completely exit BOINC and then restart it. Once activity finishes, can you cut and paste the entire contents of BOINC Manager's messages tab from the very first entry after the restart. This should be about 30-50 lines and should tell us exactly what is going on. Of course you wont need to do this if your problem turns out to be just that you had work requests disabled as I mentioned previously.

Cheers,
Gary.

Pooh Bear 27
Pooh Bear 27
Joined: 20 Mar 05
Posts: 1376
Credit: 20312671
RAC: 0

Did you download one of the

Did you download one of the optimized applications? If so, you need to stop BOINC, delete the app_info.xml file, then start up BOINC. You will get new work, then.

Ananas
Ananas
Joined: 22 Jan 05
Posts: 272
Credit: 2500681
RAC: 0

I think, the Einstein server

I think, the Einstein server sends a specific message for the case that none of the applications referred to by app_info.xml matches the WUs that the server wants to send, so it's probably not caused by this.

It might be a debit issue. If some other project has very high debits, it can stop downloads for all other projects.

@Niemz : what happens, if you pause all projects except for Einstein? (should be only CPDN I guess)
________________________

p.s.: It might be something worse, like a damaged client_state.xml or a wrong account ID. Last host contact has been 2006-06-09 and not 2006-06-21 like it's listed in Niemz' last log.

Maybe you could check the entry in account_einstein.phys.uwm.edu.xml and compare it to the account ID that you received by email. Don't post it here, it needs to be kept private - just compare them.

PiSA_Fan
PiSA_Fan
Joined: 21 Jun 06
Posts: 4
Credit: 2081296
RAC: 0

RE: Your first message has

Message 36980 in response to message 36977

Quote:

Your first message has certain information under your name at the left which gives your userID as 185089. On the website, your computer has a host# of 586023. Your second message has the same name but a new userID of 200627. Your host# is now 672428. So you have at least two separate accounts in the system.

OK. I reconnected and have the old computer (host# of 586023) again. But noe I found the former entry in the server logs (User has another host with same CPID.)

Additional here the complete log of the client:
---------
28.06.2006 18:15:03||Starting BOINC client version 5.4.9 for windows_intelx86
28.06.2006 18:15:03||libcurl/7.15.3 OpenSSL/0.9.8a zlib/1.2.3
28.06.2006 18:15:03||Data directory: C:\\Programme\\NetTools\\BOINC
28.06.2006 18:15:03||Processor: 1 AuthenticAMD AMD Athlon(tm) 64 Processor 3700+
28.06.2006 18:15:03||Memory: 2.00 GB physical, 3.85 GB virtual
28.06.2006 18:15:03||Disk: 68.68 GB total, 42.30 GB free
28.06.2006 18:15:03|BBC Climate Change Experiment|URL: http://bbc.cpdn.org/; Computer ID: 217605; location: ; project prefs: default
28.06.2006 18:15:03|Einstein@Home|URL: http://einstein.phys.uwm.edu/; Computer ID: 586023; location: work; project prefs: default
28.06.2006 18:15:03||General prefs: from BBC Climate Change Experiment (last modified 2006-06-02 12:47:25)
28.06.2006 18:15:03||General prefs: using your defaults
28.06.2006 18:15:03||Local control only allowed
28.06.2006 18:15:03||Listening on port 31416
28.06.2006 18:15:03||Suspending network activity - user is active
28.06.2006 18:19:16||Rescheduling CPU: result suspended, resumed or aborted by user
28.06.2006 18:32:21||Rescheduling CPU: result suspended, resumed or aborted by user
-----------

bye
Falko

PiSA_Fan
PiSA_Fan
Joined: 21 Jun 06
Posts: 4
Credit: 2081296
RAC: 0

RE: Did you download one of

Message 36981 in response to message 36978

Quote:
Did you download one of the optimized applications? If so, you need to stop BOINC, delete the app_info.xml file, then start up BOINC. You will get new work, then.

I couldn't find a app_info.xml in my boinc dir.

bye
Falko

J D K
J D K
Joined: 27 Aug 05
Posts: 86
Credit: 103878
RAC: 0

http://boinc-wiki.ath.cx/inde

http://boinc-wiki.ath.cx/index.php?title=Sending_scheduler_request_to_%27%28url%29%27 look here

This is from the WIKI
Note: not requesting new work or reporting results
We are in no need of work from this project.

Udo
Udo
Joined: 19 May 05
Posts: 203
Credit: 8945570
RAC: 0

RE: ... 28.06.2006

Message 36983 in response to message 36980

Quote:
...
28.06.2006 18:15:03||Disk: 68.68 GB total, 42.30 GB free
28.06.2006 18:15:03|BBC Climate Change Experiment|URL: http://bbc.cpdn.org/; ...

do you also run a WU from 'BBC Climate Change Experiment'?
Then you are possibly in 'EDF-Mode' (Earliest Deadline First, see this Wiki entry).

That means your BOINC client calculated an estimated duration for your WUs (duration_correction_factor * resource_share * time your comouter is switched on and BOINC is running * WU duration)
Especially a low resource_share for 'BBC CCE' can lead to the situation that BOINC client 'thinks' your computer will not finish that WU in time!
Therefore no new WUs from either other project are accepted.

Try to 'pause' the 'BBC CCE' WU and retry requesting E@H WUs.

Udo

PiSA_Fan
PiSA_Fan
Joined: 21 Jun 06
Posts: 4
Credit: 2081296
RAC: 0

RE: do you also run a WU

Message 36984 in response to message 36983

Quote:
do you also run a WU from 'BBC Climate Change Experiment'?
Then you are possibly in 'EDF-Mode' (Earliest Deadline First, see this Wiki entry).

Hi Udo,

I believe you are right. My Client is connected to BBC and Einstein with 50% resource shares each. The BBC WU has a Deadline at 13.05.2007 and the estimated time is 1915:28:00 h. With 8 hours calculating and 5 days a week it's almost impossible to finish in time.

In the past I run the Boinc only when no user activitys. Now I run it for ever. Lets see what happens.

Thanks
Falko

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.