Unexpected Downtime

[Darren]

Morning Guys,
As you know we have experienced a couple of 30 minute sessions of downtime recently, this is in a bid to see what works and what doesn't.

As it's a new box placed into production it hasn't had much of a break since day 1 with running more than it ever will to get the best out of it in the long run.

Now it's a process of making some changes and seeing how we go.

Email is being a real pain, despite being placed to load when the system boots it's just not wanting to work so I have to jump in there and manually start it. If at anytime you come across email not functional ie not receiving daily reports, replies etc just drop me a message on MSN and I'll get it sorted as soon as I can. Obviously email isn't an option as I won't get it.

We are doing a kernel upgrade sometime in the next week or so as the kernel we are running has multiple issue's and who knows could prevent the system from just locking up. I guess that's the thing about linux and new hardware.

Thanks for your perseverance.


Cheers
Darren

[minux]

Thanks for informing us Darren appreciated. Could be a kernel issue with email not wanting to start, can't you just cron job after a boot up say 5 mins after a server startup?

[Darren]

To update you guys as you know we went down twice yesterday, things have been going great guns for a number of weeks so we thought we had the problem licked but it's back to the drawing so to speak.

The difficulty being it can go down at anytime without any prior warning and when you get in there the system simply responds to the power button and nothing more.

Email didn't restart so that's another problem, I do recall placing a cron job in there for mail to start previously but that didn't appear to have happened either.

I have just started the mail service so a flood of emails just left and back down a 21 in the queue which are probably bounces.

Sorry about this guys, it's as frustrating for you as it is me as this simply shouldn't be happening.

[commod8]

no worries mate. thanks for letting us know

[Darren]

Everything is set for another Kernel upgrade, must be something with new hardware and linux but we can only give it a try.

It's set for tomorrow morning around 7am Adelaide time, 7:30 Eastern states and too early for the Western to be worried about. All going well it should take no more than 5 minutes assuming everything goes to plan.

[greenfoam]

I've got my money on email not starting having nothing to do with the kernel lets see tomorrow

[WAGON MAN]

I've got my money on email not starting having nothing to do with the kernel lets see tomorrow

Don't you mean yesterday 01/08/2006

[greenfoam]

:P Ok yesterday, still got nothing to do with the kernel

[Darren]

Well it's probably a little to early to say but well *touch wood* things have been great with uptime for close to 2 weeks.

One interesting thing was I was reading about as you do and came across HT (HyperThreading) causes crashes, when I went in and rebooted the machine last time it went down I switched off HT and well I think the uptime figures speak for themselves.

Lets just hope I didn't just jinx myself.