#Failed to mount the package ?

1 messages · Page 1 of 1 (latest)

stray kelp
#

Anyone else got this error while trying to update ONTAP? From 9.16.1RC1 to 9.16.1GA... in a 4 node cluster...
This happens while verifying the image... I tried to delete the image and reupload it, with no luck...
There is a KB article on this which just points to nothing other than opening a case.. I was thinking of just doing the update manually one node in each cluster at a time... should also work? (priv set diag;node image get ...) like in the "old days" 😉

proven mason
#

ah, nice, the "stale mount" bug is back. Great. I don't remember what the workaround was when this first happened in 9.13 or something, but I think back then even the node image process would fail. I mean you could always try it, it will still work (even though it's hidden behind diag now). But something else you could try is takeover/giveback of the affected nodes to see if that clears the issue

stray kelp
#

My thoughts exactly, I am currently rebooting one node at a time.. 🙂

mystic silo
#

But yeah reboots should help

stray kelp
#

the good old freebsd systemshell 😉

stray kelp
#

did the trick allright... 🙂 why does the verification always have to be run twice? Once when you start, and then you "always" have something that has a warning... you then choose to update despite of these warnings... it then does another verification before it starts the upload of the images to the nodes.... makes no sesnse to me... they should put in a timer, so if you choose to continue within 10 mins. it just starts without the extra verification... would save some time 🙂

#

....also while I am at it... why isn't it a possibility to update one node per cluster at the same time? this would also save some time... (I know it's not for everyone, but still...)

#

3hrs for an update of 4 nodes is just crazy to me 😉 this is our own system, so no consultancy fees for this time 🙂

#

Sometimes I miss the good old 7mode 😉

mystic silo
#

Unfortunately ONTAP only switches to the batch update method with clusters of 8 nodes or more.

stray kelp
#

We are not even half way yet 😉 I do not have good experience with lowering the "stability time" I had a cluster that ended up kinda broken when I was in a hurry 😉

mystic silo
#

If you regularly update your cluster (at least twice a year) and know how your environment behaves in my experience you usually can hit update and just let it run and do your regular work (or not).
As long as you've checked the requirements and the AIQ risks it's actually really stable nowadays.
No need to watch the whole process through SP console.

#

With customers after starting the ONTAP update we usually do other update stuff or just smalltalk. 😬