Actually I went back and it's not the region that is the factor... It is the image that was used to build the droplets. The Debian 7.0 images all require the kernel to be set in the control panel. Even though I've upgraded a few of those to 8.x so looks like I need to just rebuild them (thank goodness I use salt to manage them so that won't be an issue). The Debian 8.x images do not manage the kernel through the control panel. On 10/5/2015 2:43 PM, Jeremy T. Bouse wrote: > > On 10/5/2015 12:05 PM, Reco wrote: >>> 2. I don't know when 3.2.0-4 was released; I suspect the answer is yes. >>> >>> 3. many reboots; the last one earlier today. >>> >>> I note bug=767836 describes this problem, but appears closed with 3.2.0-4 >> It was closed because the problem was not in the kernel in the first >> place. It was closed because (see Message #43) VPS bootloader required >> special trickery on kernel upgrade, and that trickery was not applied. >> A classic local configuration problem (although a weird one). Thanks, >> now I know who's VPS I should never buy :) >> >> Reco >> > It is my understanding that the kernel selection is only on the > DigitalOcean droplets that haven't had the system update or were created > before the change was made. I have 2 Droplets in NYC2 that have this > kernel setting issue but are scheduled for upgrade on 10/15/15 but my > newer instances in NYC3 if you go to the Settings->Kernel like mentioned > it tells you that it is managed in the system not through the control panel. >
Attachment:
smime.p7s
Description: S/MIME Cryptographic Signature