As you probably know, the recent disclosure of serious cpu bugs created a situation in which we have to mitigate it as soon as possible.
Unfortunately, we had some work planned and were in the middle of it during this time.
We will try to do both at once to minimize downtime.
What will you see?
We will apply multiple reboots and move some containers from older nodes as well as upgrade kernels.
The XenPower installations are not compatible with the updates, the PV ones, due to a specific combination of factors. There will be an announcement regarding this. We will patch them in the meanwhile, but this can introduce instability and (maybe frequent) crashes, in this case there will be no choice but to retire this specific setup.
We hope you can understand these difficulties and our efforts to minimize the impact on your service. Please be aware we monitor intensely the situation and any crash or reboot should not last more than half an hour. If your vm is still down after 30 minutes, please login at the console and perform a manual fsck answering yes to all questions. We would also appreciate if you disloge all ISOs from the virtual CDs as this may crash your VM on reboot and cause the fs corruption I mentioned. In 99% of the cases, the fsck will solve all issues, but please do not take the risks.
Also, kernel updates are welcome, we really hope nobody runs centos 5 anymore, nor debian 6 and the like. If you do on XenPower, your VM will likely not work in the new environment, so please upgrade immediately.
Saturday, January 6, 2018
Powered by WHMCompleteSolution