I've got an HP DL 380 G6 that has been having strange intermittent issues. It all started around a year ago when it was in production. I got the 10 PM call that something was wrong, stumbled to the datacenter, and found this server in a boot loop flashing on and off. I finally got the thing to boot through a combination of breaking it down to minimum configuration, cursing, and dumb luck. I say the latter because it hasn't been consistent since.
Since that night it's been part of my test environment. It was running just fine for a number of months before I turned it off simply because I didn't need it. Now I'd like to get it on again and the problem is back. I've tried the following:
- Removing power and waiting a day.
- Minimizing the memory configuration.
- Cursing
- Following the classic HP flowchart
- Pulling the procs and reseating them both in the primary socket to determine if one is perhaps bad.
- Breaking the server down to minimum configuration.
- More creative cursing.
iLO appears to be non-functional in this state, and when I was able to get into back when it worked there were no faults recorded. The UID light doesn't respond when pressed, except when it's first plugged in. Almost as if the fault strikes several seconds after power is applied. In the past when it has turned on it did so suddenly after thorough prodding, but without touching the power button.
Anyone seen anything like this before? Is this machine doomed to sit with it's one amber eye staring into oblivion? Should I just accept that trying to resurrect a 6 year old server that's now going for sub-$200 prices on ebay maybe isn't worth it? Have I shamed it into a non-functional state with my offensive and often out-dated slurs? Will I ever run out of one-liner questions? Time will tell.
Yes, there's a process for this.
It's a weird one, and it takes a bit of patience, but please follow the steps listed at: HP ProLiant DL380 G7 servers will not POST