linux-hotplug.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Krishna Pidamale <Krishna.Pidamale@Sun.COM>
To: linux-hotplug@vger.kernel.org
Subject: Re: Question on routing hotplug events to open-ipmi interface
Date: Sat, 19 Aug 2006 01:02:34 +0000	[thread overview]
Message-ID: <44E6632A.4050201@sun.com> (raw)
In-Reply-To: <44E55918.9000508@sun.com>

THANK YOU  Kay  for your input,  let me dig udevd and more.

-Krishna


Sievers wrote:

>On Thu, 2006-08-17 at 23:07 -0700, Krishna Pidamale wrote:
>  
>
>>My system has plenty of hotpluggable SATA drives, and system runs either SLES or RHEL (older and newer distributions). Through /etc/hotplug/scsi.agent I could verify all hotplug add and remove events with DEVPATH. I would like to pass these event info to OPEN-IPMI interface, which means I will write a small C 
>>program to do this. Can I execute this program from scsi.agent? I am concerned about performance hit due to "n" instances of this executable. Since drives can be added/removed several times, it could add up many events. 
>>
>>Is this /sbin/hotplug utility meant for such operation? Any suggestions to avoid performance degradation? How can udev help here? Where can I find additional information? 
>>    
>>
>
>Depends on the size of you box. The only sane thing you can do
>with /sbin/hotplug is to disable it. :) It's simply a fork-bomb that
>kills your box when you generate a lot of events.
>
>Old SLES kernels have kernel code, to limit the number of parallel
>running /sbin/hotplug processes, standard kernels may create
>out-of-memory situations on big boxes.
>
>Newer udev versions connect with the udevd daemon to the kernel over a
>netlink socket and don't exec() a binary for every event. Also the
>created system load created by event handlers is limited by udevd.
>
>There is no way to work with /sbin/hotplug in a performance critical
>setup or on big boxes. Use a newer distro release with a recent kernel
>+udev, if you expect a lot of events. Everything else may work, but if
>your setup creates more than the usual number of events you may not even
>be able to bootup with /sbin/hotplug in use.
>
>Thanks,
>Kay
>
>  
>



-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid\x120709&bid&3057&dat\x121642
_______________________________________________
Linux-hotplug-devel mailing list  http://linux-hotplug.sourceforge.net
Linux-hotplug-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel

      parent reply	other threads:[~2006-08-19  1:02 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2006-08-18  6:07 Question on routing hotplug events to open-ipmi interface Krishna Pidamale
2006-08-18 11:28 ` Kay Sievers
2006-08-19  1:02 ` Krishna Pidamale [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=44E6632A.4050201@sun.com \
    --to=krishna.pidamale@sun.com \
    --cc=linux-hotplug@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).