Unsolved
This post is more than 5 years old
23 Posts
0
5803
August 25th, 2017 06:00
How resolve the error The patch system is busy on node ?
Hi
I have cluster isilon with 8 nodes. The OneFS Version is 8.0.0.5 and apply the package 9.3.5 and firmware 1.18 and 1.19. also apply upgrade for BMC.
The packages did not finish applying for the error
Node 2 (id:9): The patch system is busy on node id 9
Node 3 (id:10): The patch system is busy on node id 10
Node 4 (id:11): The patch system is busy on node id 11
The solve this situation execute the following commands
isi_patch ABORT
isi upgrade archive --clear
But these commands did not release the error that is still present in the nodes not allowing me to continue with the firmware upgrades
Does anyone know how to identify the processes that are running and what action to take?
Thanks in advance
Regards



Ryan_CSULB
1 Rookie
•
64 Posts
0
August 25th, 2017 10:00
Octavio, I had something similar happen a few months ago and had to open a call with support to resolve. I believe it was the same node firmware package 9.3.5 that immediately stopped on node 1 of my 4-node cluster and left me in that state (OneFS 8.0.0.2 at the time I think, now on 8.0.0.4). I don't see any specifics in my notes, but they did have to run some commands over and above what you have tried to get it to reset. Once resolved, they offered to do the install for me at a later scheduled date (and they did my OneFS 8.0.0.4 upgrade at that time as well, which was a bonus) and by that time node firmware package 10.0.1 had come out so they used that even though it seemed the only difference between 9.3.5 and 10.0.1 was extra stuff for gen6 hardware that didn't apply to me. I wonder if something is up with that 9.3.5 version?
Sorry I don't have the specifics on what they did, but they resolved it pretty quick so that might be your best option. Once resolved, I'd probably try with 10.0.1 if I were you.
LucSimard
12 Posts
0
August 25th, 2017 14:00
Octavio,
Do you have an SR (Support Case) with updated logs attached to it? Can you share the SR # Please ?
OctavioPalomino
23 Posts
0
August 26th, 2017 08:00
Hi Luc
For the momento i dont have a #SR.
I'm investigating because I happened a couple of times, I wanted to know if it happened to someone else and what action I take. I am also interested in knowing how to identify processes that are active as those that shows the error
thanks
OctavioPalomino
23 Posts
0
September 11th, 2017 14:00
I opened a case and connected an engineer. The issue is resolved by aborting the processes hanging on each node, it is not enough to eliminate them at the node connection level.
to be able to do that they are executed using the command isi_for_array -s and the command that we would like to execute.
Thanks for the answers.