Quote Originally Posted by Rota View Post
But, 2 hours is not bad downtime for an update.
Perhaps not, but the whole affair was, yet again, handled very unprofessionally. Also, I'm not sure I buy into the idea that "all servers" got a "performance upgrade" and it took 1-2 hours. There are only a few possibilities here:

1) Memory was upgraded across the board.
2) A newer (faster) router / switch was swapped into place.
3) A newer (faster) network storage device was swapped into place.

For the memory upgrade, if the "servers" are really VMs, then there would be a limited number of physical systems that needed to be taken down and have the memory installed and then booted up.

For the router upgrade, older routers (and/or switches) are taken down and new ones that have faster processors / more memory are put in their place.

Those are the only things really "safe enough" to do all at once.

One thing that makes me wonder about item 3 is the fact that there was a database error here for the forums for a bit... Could be item 2 causing that as well, with a routing table that hadn't properly been updated and thus you'd get some sort of generic error trying to connect to the database.

It would be nice if they'd elaborate on what they did. What can it hurt?