Unsolved
This post is more than 5 years old
32 Posts
0
48305
Replication partner down
I'm setting up replication on 2 EQ SAN's on different subnets. They are being routed through 2 S4810 switches. IP connectivity and routing appears to be fine as we can successfully test with other devices talking to each other across the 2 subnets. Setting up the replication partnership from SAN to SAN works fine and it passes the validation test. But when trying to replicate an actual volume, the inbound side reports no errors or warnings and gives a status of "in progress". The outbound side also reports a volume replication status of "in progress", but it also shows an overall replication status starting at yellow with message "waiting", and then finally red with message "partner down". Any ideas on what the issue might be?
Anonymous
5 Practitioner
5 Practitioner
•
274.2K Posts
0
September 9th, 2014 12:00
Great! Thanks for letting us know.
Anonymous
5 Practitioner
5 Practitioner
•
274.2K Posts
0
September 9th, 2014 12:00
Make sure you can ping from every IP address on each array to every other IP address on the remote arrays. Also make sure that port 3260 is open, that's the port replication uses to pass data.
Anonymous
5 Practitioner
5 Practitioner
•
274.2K Posts
0
September 9th, 2014 12:00
Also make sure you link isn't ignoring the do not fragment bit. Replication will fail is the packets are fragmented. If the WAN advertises that it can support jumbo frames, but not pass them this too can cause this problem.
You can disable jumbo frames for replication with this CLI command.
GrpName>support repl-use-jumbos no
tickermcse76
32 Posts
0
September 9th, 2014 12:00
We figured this out; some of the source SAN's in the group didn't have the appropriate gateway configured. They were previously not configured to route on the iSCSI side.