We have a web app built on Jboss that has been in production for several years; for the last 18 months it has never gone down. However, in the last two days, the jvm has crashe
Are the crash reports always the same? Do you have the stack traces? i.e. the bit in the crash dump headed by
--------------- P R O C E S S ---------------
It's a v vague dump tbh. One possibility is dodgy RAM as it seems to have died while at a safepoint reached in order to collect garbage. Have you done any hardware diagnostics on the box?
I agree with Matt, it does sound like hardware problems. I've seen your three crash reports and they are all failing in different parts. One, mostly when executing VM code. The second, in native code. And the third (in the question you should have not opened, but added here), in standard Java code.
Of course, if you have several production machines and the issue is happening in different ones, then I would seriously consider to buy subscriptions and get real support.