Posts by Richard Haselgrove

1) Message boards : Technical News : Gravitational Wave search S6Bucket Follow-up #1 launched (Message 135251)
Posted 19 hours ago by Richard Haselgrove
Hallo BM!
Are these new apps running for test with high priority? I hope so.

Kind regards and happy crunching
Martin

That's not something the project can control. 'High Priority' is set by your local client if it needs to gueue-jump the task in order to meet the deadline.
2) Message boards : Cruncher's Corner : CUDA 6.5? (Message 135136)
Posted 2 days ago by Richard Haselgrove
Cuda 6.5 would definitively exclude all compute capability 1.x cards - not only the remaining legacy 8xxx, 9xxx, and 2xx ranges, but also some Quadro FX and even early Tesla models.
3) Message boards : Problems and Bug Reports : Einstein, cc_config exclusions and Work unit limits (Message 135050)
Posted 4 days ago by Richard Haselgrove
It works fine with them, but you'll need to catch the server logs at the right time - i.e. when the request you're interested in has just (most recently) happened. We don't have access to a full history, just the most recent.
4) Message boards : Problems and Bug Reports : Einstein, cc_config exclusions and Work unit limits (Message 135044)
Posted 4 days ago by Richard Haselgrove
So,

Having sufficient time on my hands to test some things here, I've noticed an issue with receiving work units.

I've set up my cc_config to maximize my gpus while running Einstein. What I've found is that after the initial request for work units, the server will not provide my computer with sufficient work units after all work is completed.

Example, I request a 5 days worth of work and receive it initially. But after I have crunched all that work and only have 2-3 left, the server will deny my computer any more work. It is only when I am completely out will it parsal out 2 or 3 work units rather than a 5 day allotment.

There are no other projects requesting work.

Resetting the project has no effect.

The only way I have found to resolve this is it remove the project completely and reinstall it.

Ideas?

As always, read the logs.

You have two sets available here:

1) <work_fetch_debug> in your local BOINC client, to see what work you're requesting and why (or why not)
2) Server logs, to see what request was received and how it was dealt with. At the moment, both http://einstein.phys.uwm.edu/host_sched_logs/10267/10267709 and http://einstein.phys.uwm.edu/host_sched_logs/11681/11681771 are indicating that no work was requested on the last contact. So look to your client first - sending no work when none is requested doesn't amount to a "denial", in my book.
5) Message boards : Problems and Bug Reports : All Einstein cpu tasks stuck at 98% (Message 134800)
Posted 9 days ago by Richard Haselgrove
I forgot to comment on the CPU support needed in my previous post.

Archae86's observations is still true, the Intel GPU app needs a full CPU core to support it or the run time of the tasks really blows up, it's on the scale of taking hours compared to about 15 min per task. Checking in windows task manager the currently running Intel GPU tasks has an elapsed time of 10 min but has only used 21 seconds of CPU time, mostly the process reports 0% CPU usage with the occasional 1-2% blip.

Also confirmed, in response to Archae86's post. Tried my Haswell i5 HD4600 (driver version 10.18.10.3621) with all four cores loaded: runtime increased ~7x (on the basis of taking seven minutes to reach 10%, against the normal 11 minutes for the full task).
6) Message boards : Problems and Bug Reports : two GPUs seen but one is "ignored by config" (Message 134664)
Posted 13 days ago by Richard Haselgrove
In case of doubt, always keep a link to client configuration handy.
7) Message boards : Cruncher's Corner : Maxwell 2 (Message 134663)
Posted 13 days ago by Richard Haselgrove
'Recent' version of BOINC is a relative term. According to client configuration, <use_all_gpus> has been available since v6.6.25
8) Message boards : Problems and Bug Reports : How to release a work-unit stuck in "upload-pending"? (Message 134603)
Posted 14 days ago by Richard Haselgrove
I've got a BRP4 upload stuck at the moment, and getting the same

07/11/2014 16:29:24 | Einstein@Home | [http] [ID#35] Received header from server: HTTP/1.1 502 Bad Gateway

response from the server that I reported before.
9) Message boards : Cruncher's Corner : A lot of invalid results (opencl-intel_gpu) (Message 134464)
Posted 20 days ago by Richard Haselgrove
oops sorry.
I tried the previous version earlier, which was even worse.
I guess the version has to match the one the wingman is using.

No, it doesn't - it just has to be compatible with the project application(s), and free of Intel bugs.

I've found that the best version, both here and at SETI, is 10.18.10.3621, released May 21 2014. I haven't checked to see how easy it is to retrieve that version from Intel's download site, but I have the files (64-bit Windows) here if anyone can host them.
10) Message boards : Problems and Bug Reports : How to release a work-unit stuck in "upload-pending"? (Message 134380)
Posted 24 days ago by Richard Haselgrove
Found it.

26-Oct-2014 16:16:04 [Einstein@Home] [http] [ID#11127] Info: About to connect() to einstein4.aei.uni-hannover.de port 80 (#11588)
26-Oct-2014 16:16:04 [Einstein@Home] [http] [ID#11127] Info: Trying 130.75.116.34...
26-Oct-2014 16:16:04 [Einstein@Home] [http] [ID#11127] Info: Adding handle: conn: 0x318c5c0
26-Oct-2014 16:16:04 [Einstein@Home] [http] [ID#11127] Info: Adding handle: send: 0
26-Oct-2014 16:16:04 [Einstein@Home] [http] [ID#11127] Info: Adding handle: recv: 0
26-Oct-2014 16:16:04 [Einstein@Home] [http] [ID#11127] Info: Curl_addHandleToPipeline: length: 1
26-Oct-2014 16:16:04 [Einstein@Home] [http] [ID#11127] Info: - Conn 11588 (0x318c5c0) send_pipe: 1, recv_pipe: 0
26-Oct-2014 16:16:05 [Einstein@Home] [http] [ID#11127] Info: Connected to einstein4.aei.uni-hannover.de (130.75.116.34) port 80 (#11588)
26-Oct-2014 16:16:05 [Einstein@Home] [http] [ID#11127] Sent header to server: POST /EinsteinAtHome/cgi-bin/file_upload_handler HTTP/1.1
26-Oct-2014 16:16:05 [Einstein@Home] [http] [ID#11127] Sent header to server: User-Agent: BOINC client (windows_x86_64 7.4.22)
26-Oct-2014 16:16:05 [Einstein@Home] [http] [ID#11127] Sent header to server: Host: einstein4.aei.uni-hannover.de
26-Oct-2014 16:16:05 [Einstein@Home] [http] [ID#11127] Sent header to server: Accept: */*
26-Oct-2014 16:16:05 [Einstein@Home] [http] [ID#11127] Sent header to server: Accept-Encoding: deflate, gzip
26-Oct-2014 16:16:05 [Einstein@Home] [http] [ID#11127] Sent header to server: Content-Type: application/x-www-form-urlencoded
26-Oct-2014 16:16:05 [Einstein@Home] [http] [ID#11127] Sent header to server: Accept-Language: en_GB
26-Oct-2014 16:16:05 [Einstein@Home] [http] [ID#11127] Sent header to server: Content-Length: 4545
26-Oct-2014 16:16:05 [Einstein@Home] [http] [ID#11127] Sent header to server: Expect: 100-continue
26-Oct-2014 16:16:05 [Einstein@Home] [http] [ID#11127] Sent header to server:
26-Oct-2014 16:16:05 [Einstein@Home] [http] [ID#11127] Received header from server: HTTP/1.1 100 Continue
26-Oct-2014 16:16:05 [Einstein@Home] [http] [ID#11127] Received header from server: HTTP/1.1 502 Bad Gateway
26-Oct-2014 16:16:05 [Einstein@Home] [http] [ID#11127] Info: Server nginx/1.2.1 is not blacklisted
26-Oct-2014 16:16:05 [Einstein@Home] [http] [ID#11127] Received header from server: Server: nginx/1.2.1
26-Oct-2014 16:16:05 [Einstein@Home] [http] [ID#11127] Received header from server: Date: Sun, 26 Oct 2014 16:12:59 GMT
26-Oct-2014 16:16:05 [Einstein@Home] [http] [ID#11127] Received header from server: Content-Type: text/html
26-Oct-2014 16:16:05 [Einstein@Home] [http] [ID#11127] Received header from server: Content-Length: 172
26-Oct-2014 16:16:05 [Einstein@Home] [http] [ID#11127] Received header from server: Connection: keep-alive
26-Oct-2014 16:16:05 [Einstein@Home] [http] [ID#11127] Info: HTTP error before end of send, stop sending
26-Oct-2014 16:16:05 [Einstein@Home] [http] [ID#11127] Received header from server:
26-Oct-2014 16:16:05 [Einstein@Home] [http] [ID#11127] Info: Closing connection 11588
26-Oct-2014 16:16:06 [Einstein@Home] Temporarily failed upload of p2030.20131212.G177.70+00.11.C.b4s0g0.00000_3214_1_0: transient HTTP error


Next 10

Home · Your account · Message boards

This material is based upon work supported by the National Science Foundation (NSF) under Grants PHY-1104902, PHY-1104617 and PHY-1105572 and by the Max Planck Gesellschaft (MPG). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the investigators and do not necessarily reflect the views of the NSF or the MPG.

Copyright © 2014 Bruce Allen