dsoltesz's Posts

Thanks, but those patches are from 2009 and I have had them applied to my system for a long time now and still see the same issues.  I'm even running 4.1 now and VMs still do the same things.  Like I said in an earlier post, it is just to OS touching all the memory.  After a couple of minutes the usage settles provided a VM has enough memory. Doug

Lars, I can tell you what I have done in my environment.  I'm not saying that this is the best answer, but it works for us. 1.  I have edited the alarm definition for Guest Memory Usage under datacenter, alarms, definition.      It is set to Warn if condition is above 85% for 15 minutes, or Alert if condition is above 95% for 10 minutes. This way when a VM boots, it touches 100% of the memory, but over the first 10 minutes some of the memory is not used.  As long as 5% isn't not used after 10 minutes, the Alert trigger will not hit.  Then in the next 5 minutes, the amount of active memory should fall another 10% so that the warn flag should not hit.  Now if either of these conditions do hit, and stay on, I know that I probably need to give that VM more memory, as it is actively using the memory I have given it. Most VMs will have a sweet spot.  I have a SQL server running Dynamics, that was given 4GB and was constantlly alerting.  I boosted it to 6GB, and it runs better and no alerts.  I'm sure the sweet spot was lower than 6GB, but if memory becomes an issue in the environment, vSphere has memory ballooning and memory compression techniques for pulling back unused memory. Doug PS.  To help you set the tresholds, try opening the advanced performance tab for a VM.  Switch it to memory and reboot the VM.  You will see the active memory spike (Red line in my graphs).  Then after a few minutes you will see it start to settle down.  For me this occurs in about 10 minutes, thus the tresholds I set above.

Lars, If you open up Windows Task Manager you may see that the Active Memory as represented by Microsoft does not agree with vCenter. This can be for a few reasons. 1.  After a reboot/startup event Windows touches all of the memory, triggering the vcenter alert event.  The event should clear after around 20 minutes when vCenter polls what memory has changed/touched since the last poll. 2.  The amount of memory listed as used in the graph Microsoft gives does not include the system cache.  It would be foolish for any OS not to take advantage of all the memory for both user applications and system cache.  This will lower the number of SAN/storage hits, which is far slower than memory.  To see how much memory is being used add in the memory listed as cached (also in task manager) to the number presented in the task manager graph.  Or in server 2008 just check out the Line that reads Free. 3.  After awhile on a server that is not active, the active memory displayed in vmware will be much lower than the active memory in Microsoft.  Microsoft does not unload items from memory if they are not being used, and there is spare memory available.  This is just more caching.  VMware lists active memory as memory that is changed during a specific time period, which on an inactive server will be low. Doug