Used wrong flag './patcher.py -p' on a pool master....

[Keithsc]

Hi, So I accidentally used the wrong option './patcher.py -p' on one of our XenServer pool masters. The patches were applied so I reboot all the hosts in the pool and then realised that only the pool mater had been patched.

If I try to rerun the script on the pool master I get "No Patches Required. System is up to date." and if I run it on one of the un-patched hosts in the pool the script just waits at "Internal Upload..."

I am manually patching the pool using XenCenter :( and should be finished by Christmas.

I guess this isn't realy a bug ? but for anyone in the situation in the future is there work around ? Can you force patches ?

[dalgibbard]

It's an interesting point for sure. I think, given my lack of a xenserver environment for use/testing, and that we've not seen this issue before now, it's enough of an edge case to be ignored, for now at least. I do feel for you though! I guess the only side option is to break the pool to individual nodes, patch, and re-pool... But that's not necessarily a straightforward option in prod either!