[RC5] Key rate halfing on OS/2 client

Jeffrey Doolittle jdoolittle at ibm.net
Wed Nov 26 21:25:04 EST 1997

On Wed, 26 Nov 97 17:36:02 -0500, Colin L. Hildinger wrote:

>On Wed, 26 Nov 97 17:20:43 +0500, Jeffrey Doolittle wrote:
>>On 26 Nov 97 16:14:20 +0000, SCRUTE.UK.ORACLE.COM wrote:
>>>  I've discoverd that the keyrate on one of my OS/2 clients has halfed. It 
>>>happened on 14th November around lunchtime and it hasn't got back up to
>>>yet. Pulse is only regestinging about 1/2 instead of maxing out. I can't 
>>>remember if I did anything then on this PC, but I don't think so. Anyone got
>>>a clue what this could be ?  
>>I would be curious as to how long the clients have been running.  I've
>>got six computers running the OS/2 client without any problems, the
>>longest at the moment is a web server which has been operational for
>>312 hours the last time I checked.  
>Well, I ASSUME he's not running pulse all the time.  If you run pulse
>and have the client set to nice=0 you will rob 50% of the cycles (pulse
>basically tries to use up all the free idle 0 slices to determine what
>percent of your cpu time you're actually using).

I was under the assumption that pulse was only being used to determine
what could be robbing critical CPU cycles for processing keys.  If he
is indeed using pulse full time then you are correct.  I haven't tried
PM Patrol (see BMT Micro) with RC5-64 but on uni-processor servers it
works great and hardly uses any cycles (I've had problems with PM
Patrol on dual processor machines).

Jeffrey Doolittle
jdoolittle at ibm.net
Send email w/subject 'REQUEST PGP KEY' for my PGP public key.

To unsubcribe, send 'unsubscribe rc5' to majordomo at llamas.net
rc5-digest subscribers replace rc5 with rc5-digest

More information about the rc5 mailing list