Hi Dave, > Yes, but note they're computed values, while the time > values are > measured. The measured values for resource usages are > traditionally > in units like KB/S or IOPS, and they're hard to turn > into utilization > numbers. That leaves the poor sysadmin faced with coming > up > for a value of something that correlates to 100% > utilization.
One example here is that folks usually measure just the CPU utilisation as such and they make decisions based on that measurement alone, which is bogus. Example: say you have a DB server which burns CPU around 70-80%. Which might be simple ok, but sometimes they rush and simple ask their managers to buy another machine and another machine... It is good to have around the OS metrics and to relate them with u workload: tps, iops etc - This is a very often scenario, even in big DataCenters: you could have a 90% busy machine doing work and people are unhappy for that :) They look only to the CPU % and they simple dont measure the saturation or how much work/clients are waiting to be served. 90% might be just fine , finally thats why you got the box - to put it to work for you. > Conversely, Guerrilla CP is about doing a good job under > nasty time constraints, and I see a huge amount of value in > his analytic model of the utilization curve, not > least because it predicts the drop-off after > "100%" I found the book and course very valuable and easy applicable: how can you survive these crazy times when you got a very short time window for capacity planning and management is pushing and pushing... Thanks, Stefan _______________________________________________ perf-discuss mailing list perf-discuss@opensolaris.org
