1) Message boards : BOINC Manager : BOINC slowdown on Mac OSX (Message 2713) Posted 23 Jan 2006 by Jonathan Jeckell Post: Ok, no connection problems with the SETI project today, yet it happened again...twice. I also failed to mention earlier that it used to show up as "Kernel Task" like one of the earlier posters, but lately the thread has reported as SETI, but again, 40% of the available processor for no sign of work getting done, while another project (Einstein, Rosetta, or Predictor) struggles with the remaining 40%. Could BOINC be mistaking my machine for a dual processor machine and scheduling SETI and another project to both run on the same processor simultaneously?

2) Message boards : BOINC Manager : BOINC slowdown on Mac OSX (Message 2708) Posted 22 Jan 2006 by Jonathan Jeckell Post: I have a hypothesis for what caused my problem and I was wondering if any of the rest of you who had a similar problem can corroborate or deny. The times when SETI seized up seems to correlate with dates when SETI was having connection problems (like over the MLK weekend (13-16 Jan 06) outage and subsequent work to restore it. My STERRGUI.txt log has a long line of repeats of this message, but it doesn't say SETI specifically, nor do I see a date stamp that goes with it: init_poll: get_socket_error(): 61 STERR for the SETI slot is blank now, so I don't know what was in there before--if SETI did report an error to this file it was wiped out by the last restart of BOINC manager.

3) Message boards : BOINC Manager : BOINC slowdown on Mac OSX (Message 2704) Posted 22 Jan 2006 by Jonathan Jeckell Post: I have the same problem, but mine is a little different. I have a single processor G5 and most of the time that SETI runs, it will not let go of the processor when it's time for other processes to run (i.e. Einstein fires up, but SETI continues to run). I wouldn't care, except that while SETI is taking up ~40% of the processor, it's not getting any work done--it's just eating processor cycles. This leaves the other project which is supposed to be running only about 40% of the processor for it to do actual work. I checked STERR for BOINC and for SETI, but I didn't see anything jump out at me that was a problem. It's been happening for at least the last week or so for sure, so I thought it might be a server communications issue causing it to hang. It's always SETI that causes the problem; none of the other projects do this.

---

Copyright © 2024 University of California.
Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.