The hardware on mail1 is going to be changed at 11 am EDT?
Well, I guess I should start telling customers they can expect to be without email until sometime Wednesday. How many hardware changes have gone off without a hitch? And to do it during the workday for so many people. But I’m sure you have good reasons to do it then, right?
Good Oyster, we had to do this when we could. We will try to keep the outage to a minimum. This was the first scheduled opportunity to us to update the server.
Well, so far so good! Looks like this one DID go off without a hitch!
Thanks for the good work!
yes, it went as quickly as possible, and load is already coming down from gathered mail at the time it was off.
Just as a FYI, the problem stemmed from an issue that the qmail can not use multi threading, so it would exhaust all of one cpu, and leave the others doing nothing, and create timeouts.
So adding more power did no good in the case of qmail in term of adding more cpus, it just needed more Ghz per core, because it could only use the first one.