[prev in list] [next in list] [prev in thread] [next in thread] 

List:       evms-devel
Subject:    [Evms-devel] RE: [Evms-cluster] A bug in HA plugin
From:       "Thomas Guyot-Sionnest" <Thomas () zango ! com>
Date:       2005-12-22 21:37:10
Message-ID: E345C809C68668438936E25DB7EBF7FF8BFA9A () seaex01 ! 180solutions ! com
[Download RAW message or body]

Hi Steve,

This looks much like what I was seeing. Do you think we may have hit the
same bug? Should I apply your corrolator_race patch, and should I apply the
response_count patch too?

Thanks,

Thomas Guyot-Sionnest,
Administrateur de systèmes
Tél: (514) 842-7054
Fax: (514) 221-3395
Courriel: thomas@zango.com  

> -----Original Message-----
> From: evms-cluster-admin@lists.sourceforge.net [mailto:evms-cluster-
> admin@lists.sourceforge.net] On Behalf Of Steve Dobbelstein
> Sent: December 14, 2005 15:39
> To: Shuqing Yang
> Cc: evms-cluster@lists.sourceforge.net; evms-devel@lists.sourceforge.net
> Subject: Re: [Evms-cluster] A bug in HA plugin
> 
> Shuqing Yang <sqyang@gmail.com> wrote on 12/14/2005 01:47:28 AM:
> 
> > Hi all,
> >
> > I find evms may hang  occasionally when I use evms to execute remote
> > operation. Log indicate local system is waiting for response while
> > remote system is waiting for command. So  they all wait infinitely. I
> > debuged and found some reason about this.
> >
> > In function ece_send_msg (file ece.c, 2.5.4 edition), allocate
> > corrolator first( line 1954), then call frag_send_msg_wait to send
> > message(line 1964), if success, fill ecemsg->corrolator with real
> > corrolator( line 1972). If message's response comes back after calling
> > frag_send_msg_wait but before filling ecemsg->corrolator, this
> > response message will not find its matched command message. So problem
> > emerged.
> >
> > Attachment is a patch file. A little ugly but effective.
> 
> Hi, Shuqing.
> 
> Thanks a lot for finding that race condition!  And for proposing a fix.  I
> agree with your method for fixing the problem. I made a few tweaks to get
> rid of extra variables.  Here is my version.
> 
> (See attached file: corrolator_race.patch)
> 
> It works on my test machine, but then I'm not sure I'm hitting the race
> condition.  Could you look over my patch and try it out on your machine to
> make sure my tweaks didn't break the work you did?  If you say it looks
> good I'll check it in to CVS.
> 
> Thanks,
> Steve D.

["smime.p7s" (application/x-pkcs7-signature)]
-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems?  Stop!  Download the new AJAX search engine that makes
searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
http://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click
_______________________________________________
Evms-devel mailing list
Evms-devel@lists.sourceforge.net
To subscribe/unsubscribe, please visit:
https://lists.sourceforge.net/lists/listinfo/evms-devel

[prev in list] [next in list] [prev in thread] [next in thread]