Last Thur afternoon my IFL's went to nearly 100% for extended periods. O ne linux guest running three WebSphere app servers was the culprit. After
determining the cause to be the JVM of one the servers causing linux to swap the JVM, we increased the size and restarted server. Things are fin d again. Monday AM all day same situation different server CPU at or near 100% (for 2 IFL's), WebSphere Admins research and determine the JVM need s increased in another server and the guest needs more Memory. I give the guest 700M to a total size of 4G. Log machine off and back on. Since th en this guest runs in the 40-50% range vs the 90-80%. What I am unable to understand is why did one linux guest 'swapping' cause the IFL to be totally consumed? NOTE: VM worked great no other guests were effectted to the point user's complained.
