Guys,
I'm seeing an issue which I'm not sure how to solve.
I'm testing out a rolling upgrade with reboot. When SMF gets to
the node with the proxy/proxied on it, it issues the AMF node lock which
fails due to the following:
Jul 14 20:05:55 linux osafamfnd[12884]: NO
'safComp=Proxied,safSu=Management-SU2,safSg=Management-2N,safApp=ManagementApp'
faulted due to 'prxRegTimeout' : Recovery is 'suFailover'
I've traced this down to the fact that the AMF node lock is
quiescing the proxy first, which causes the proxy to unregister the
proxied comp. Now the proxied comp doesn't get quiesced. 30 seconds
later AMF tries to clean up the proxied, and the rest of the SU it is in.
This is clearly not ideal. We shouldn't be waiting 30 seconds to
do a cleanup for an AMF node lock.
I'd like to solve this in a similar way to how we did the opensafd
shutdown (by waiting for all the proxied components of the proxy to
terminate before terminating the proxy), but as far as I can tell amfnd
doesn't know that a node lock is going on. "admin_state" in the cb is
never changed.
What is the best way to handle this case? Adding support for
"admin_state" so that amfnd knows that a node lock is in progress?
Alex
------------------------------------------------------------------------------
Want fast and easy access to all the code in your enterprise? Index and
search up to 200,000 lines of code with a free copy of Black Duck
Code Sight - the same software that powers the world's largest code
search on Ohloh, the Black Duck Open Hub! Try it now.
http://p.sf.net/sfu/bds
_______________________________________________
Opensaf-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/opensaf-devel
