We are having a lot of performance problems with our baselines. Since we have so many patch groups (@30 with different day and time/windows) in which we deploy a baseline to each we notice the performance impact in response time.
One question I have is, why do machines which are not targeted by the action of a baseline try to translate the full relevance of that baseline?
Example: 3 systems A, B, C, all in the same subscribed sites. As regular user of the console, we deploy a baseline with 50 fixlets to system A only, using the targeting in the action. System B and system C, enough though they are not targeted will try and evaluate all 50 fixlets in that baseline even though those machines are not targeted. The baseline relevance is set to the default of “true”. We have proved this out with the usage profiler.
Shouldn’t the client look at the targeted relevance and once it notices the machine is not targeted, discard the action and not process the other 50 relevance statements?
Yes… although some of the older BigFix Client would try to evaluate the Fixlets anyway (which was a waste and why we optimized that behavior away)… What version are you agents using?
We are on version 7.1.1.315 with all of our clients, relays and master server, which is the latest I believe for the client and the server, although the relay has a newer version. We also have the 7.0 mime setting configuration on the server.
When we deploy multiple baselines, our computers go gray because they stop reporting within a reasonable amount of time. Running the usage profile, we can see computers that were not targeted with a baseline, evaluating fixlets from that baseline.
Well… that is the right version that is supposed to be much more efficient… are you saying you have usePre70ClientCompatibleMIME=false (which means that you are using the newer and more efficient MIME format)?