[Mldonkey-bugs] [bugs #8716] one cpu spikes to 99% after 10 hours.

[spiralvoice]

This mail is an automated notification from the bugs tracker
 of the project: mldonkey, a multi-networks file-sharing client.

/**************************************************************************/
[bugs #8716] Latest Modifications:

Changes by: 
                spiralvoice <address@hidden>
'Date: 
                Thu 07/22/04 at 23:50 (Europe/Berlin)

            What     | Removed                   | Added
---------------------------------------------------------------------------
          Resolution | None                      | Wont Fix
              Status | Open                      | Closed


------------------ Additional Follow-up Comments ----------------------------
As this bug seems to be hardware related it is closed.






/**************************************************************************/
[bugs #8716] Full Item Snapshot:

URL: <http://savannah.nongnu.org/bugs/?func=detailitem&item_id=8716>
Project: mldonkey, a multi-networks file-sharing client
Submitted by: 0
On: Wed 04/28/04 at 04:28

Category:  Core
Severity:  7 - Major
Item Group:  Program malfunction
Resolution:  Wont Fix
Assigned to:  None
Status:  Closed
Release:  2-5-18
Release:  started in 2.5.12
Platform Version:  Linux
Binaries Origin:  Downloaded from third-party page
CPU type:  Intel x86


Summary:  one cpu spikes to 99% after 10 hours.

Original Submission:  dual P2 266 machine with 256MB ram.

I started having this problem around 2.5.12.  i have tried SpiralVoice's cores 
as well as building my own cvs cores.  all give the same results.  this last 
time i used the 2.5-18 core from Spiralvoice's page and I had disabled all 
networks except eDonkey some 7 hours before it crashed.

everything seems to run fine and downloads are fast, but after about 10 hours 
of running, mlnet cpu useage will spike to 99% and i can no longer log into the 
core.  I have to kill-9 it. 

i have made the buginfo and included the logfile in the archive 
(mld.log.crash).  i can get more info if required and i'm more than happy to 
help any way i can.

I've toyed with various options settings, but, like i said, everything else 
seems to be fine (mem useage, etc...).  and i have trouble figureing out 
exactly when it's going to crash.

Follow-up Comments
------------------


-------------------------------------------------------
Date: Thu 07/22/04 at 23:50         By: spiralvoice
As this bug seems to be hardware related it is closed.

-------------------------------------------------------
Date: Thu 05/06/04 at 13:32         By: fatsheldon
two things happened.  i changed to another machine (almost identical - p2 300 
(x2) with 320MB ram, same model MB but this one's bios chip isn't dead),  and i 
started using the 2.5.20 spiralvoice core.

good news and bad news.

bad news is that i can't get the strace for you.

good news is that i can't get the strace because the thing won't die.  In 
addition to that it's faster than ever...  i've had to go looking for stuff to 
download just to keep it busy!

if it crashes i'll follow up with an strace first thing, but i'm keeping my 
fingers crossed.  -it's been running for 15 hours and downloaded close to 1GB 
of files.  i put the machine under heavy load, and put the local network under 
heavy load.  the net connection even went out for a little while-  no problems 
yet!!

-------------------------------------------------------
Date: Tue 05/04/04 at 19:14         By: spiralvoice
I quickly looked at the logfiles but could not see something obvious. Could you 
send a strace log of the mlnet process when it blocks the CPU? Maybe this 
problem is SMP-related?

-------------------------------------------------------
Date: Tue 05/04/04 at 19:09         By: spiralvoice
The data loss should be repaired in current CVS versions:
2004/05/02: Fabrice
  - Fixed bug causing lost of partially downloaded chunks.

-------------------------------------------------------
Date: Wed 04/28/04 at 04:50         By: fatsheldon
one more thing i just noticed (aside from the fact that i left the post as a 
"Compilation problem" when it is obviously not).

when i restarted the core i had lost considerable data on two of the 4 
downloading files.  roughly 10MB on one and about 60MB on another. (at least it 
restarts now though)

the other two files seemed the same.

-------------------------------------------------------
Date: Wed 04/28/04 at 04:39         By: fatsheldon
sorry, i had forgot to login... stupid...

also, forgot to mention that with the 2.5.16f and 2.5.16k cores after this 
problem happened once i could no longer even start the core again without it 
doing the same thing right away.  With 2.5.18 however i can start it up again 
without trouble.  so, at least that's something.






File Attachments
-------------------

-------------------------------------------------------
Date: Wed 04/28/04 at 04:28  Name: buginfo.tar.gz  Size: 244.89KB   By: None
buginfo + logfile
http://savannah.nongnu.org/bugs/download.php?item_id=8716&amp;item_file_id=1244






For detailed info, follow this link:
<http://savannah.nongnu.org/bugs/?func=detailitem&item_id=8716>

_______________________________________________
  Message sent via/by Savannah
  http://savannah.nongnu.org/