2012-10-17 02:48 AM
I have the following setup:
Active Active properly setup.
SP properly setup & I can use it.
e0M interfaces setup & they can ping each other.
e0M from the same subnet as the rest of interfaces or the SP subnet is unreachable eventhough I had the switch port as trunk & I have a router properly configured.
Now when I configure the cf hardware assist, it shows me active but testing fails & I see exactly the below:
NAS-THQ-1A> cf hw_assist test
cf hw_assist Error: No response from partner(NAS-THQ-1B), timed out.
NAS-THQ-1B> cf hw_assist test
cf hw_assist Error: No response from partner(NAS-THQ-1A), timed out.
NAS-THQ-1A> cf hw_assist status
Local Node(NAS-THQ-1A) Status:
Active: NAS-THQ-1A monitoring alerts from partner(NAS-THQ-1B)
port 4444 IP address x.y.z.121
Missed keep alive alert from partner(NAS-THQ-1B).
No keep alive alert received from partner since last boot.
Partner Node(NAS-THQ-1B) Status:
Active: NAS-THQ-1B monitoring alerts from partner(NAS-THQ-1A)
port 4444 IP address x.y.z.122
The ip addresses used are the e0M, the moment I switch to any other interface, everything is ok. I searched the forum upside down, & documents & I understood that for it to work, you need IP reachability between SP & the e0M which am not finding a way to establish.
I tried doing SP setup again. Still no progress. I know that the best practice is to use the e0M & it should be up for the feature to work, but you can literally use any IP. I need the e0M.
Kindly does anyone know how to solve this?
2012-10-17 10:29 AM
Just to clarify, if e0M is not in it's own dedicated subnet for strictly management traffic, you're better off disabling it.
Also, if the e0M interface lives in the same subnet as your default gateway it can cause problems.
Setting CF assist isn't so much a best practice as it is the default interface automatically used if it's configured. HW assist will work perfectly on any ip it generates as long as the SP is functioning properly.
I would fix the e0M issues and test it again. If it still does not work can you send us your RC file?
Hope this helps,
2012-10-17 11:04 AM
Is your e0M IP address on the same subnet as your RLM/SP? (x.y.z.123/4 for example), it's one cable so you can't use any IP.
Check options interface.blocked.mgmt_data_traffic.
As stated check your route, do a traceroute from the controller to the IP that is trying to access your e0M IP address, make sure it routes over e0M IP address.