Modify "write_wakeup_threshold" in RHEL 5.6
Joshua Newton
newton.joshua.x at gmail.com
Sat Dec 1 23:48:43 GMT 2012
It sort of depends. If your application needs more than ~32Kbps of entropy,
then you're not missing anything, except perhaps the specifications page. ;)
If you do need more entropy, I'm finding haveged (
http://www.issihosts.com/haveged/ ) to be a rather good replacement for the
Entropy Key. Depending how much entropy you need, and what testing options
you use when invoking haveged, you might see some CPU hit. You'll probably
want to adjust the wakeup for haveged, too. It's expressed as a low
watermark for kernel.random.entropy_avail -- with nearly everyone's kernel
fixed at a maximum of 4096 bytes, I set mine to 3072.
Complete command line:
/usr/sbin/haveged -w 3072 -o tbacab
If you want RPMs, there are some out there. I've patched up one of the old
Fedora (13?) spec files to work on RHEL / CentOS 5.x and 6.x, and provide
the latest haveged (1.5). I can provide an SRPM when I get back to work on
Monday, if you're interested. Contact me directly.
On Tue, Nov 27, 2012 at 1:39 PM, Dhabuwala, Nikhil
<NDhabuwala at medplus.com>wrote:
> Our application makes a pretty heavy use of “random” and without the
> Key, we used to experience a significant performance bottleneck.
> Even with the key, we see the “entropy_avail” is always pretty low.
> Running “watch -n 0.1 cat /proc/sys/kernel/random/entropy_avail” we see the
> numbers jump all over the place, generally staying in the 100s range,
> reaching a max of 4096 and going as low as 128 (occasionally lower)
>
> We modified the “write-wakeup_threshold” from 128 to 1024. Unfortunately,
> this does not seem to have made any difference.
> Is there anything else we are missing?
> We are running RHEL 5.7.
> Any suggestions appreciated.
>
> Thanks!
>
> *Nik Dhabuwala*
>
>
>
>
>
>
>
> Confidentiality Notice: The information contained in this electronic
> transmission is confidential and may be legally privileged. It is intended
> only for the addressee(s) named above. If you are not an intended
> recipient, be aware that any disclosure, copying, distribution or use of
> the information contained in this transmission is prohibited and may be
> unlawful. If you have received this transmission in error, please notify us
> by telephone (513) 229-5500 or by email (postmaster at MedPlus.com). After
> replying, please erase it from your computer system.
>
> _______________________________________________
> EntropyKey-users mailing list
> EntropyKey-users at lists.simtec.co.uk
> http://lists.simtec.co.uk/cgi-bin/mailman/listinfo/entropykey-users
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.simtec.co.uk/pipermail/entropykey-users/attachments/20121201/822da0af/attachment.html>
More information about the EntropyKey-users
mailing list