That's correct and in fact a better idea. But no configuration to be done on the Data Mover end (DM3 is a standby one and gets the configuration of server_2 when failed over to) - only to fix the network connection and etherchannel configuration on the switch.
I think I might have answererd my own question with an alternate solution. I can configure the standby dm03 with the correct network cabling and etherchannel configuration and simply failover the prod dm02. then reconfigure the prod dm02.
Apparently - you can simply unplug the cable from cge1 and connect the new cable. Since cge0 and cge 1 are already trunked on the Data mover - it does not require any configuration change on the NAS end - only configure the channel on the switch ports.
However, few things to consider -
1) I am assuming that currently only cge0 is being used (as the valid connection in the trunk) for all user connections - however, you may confirm that looking at the network switch port activity and throughput. The port to whcih cge 1 is currently connected should not any traffic.
2) when you configure Etherchannel on the switch - it may cause a momentary loss of connectivity for cge0. You may use portfast feature on these ports to minimise the time. UNIX clients will not be affected as NFS will keep on retrying the connection - but windows client may loose their session and have to reconnect to the shares on NAS.
nandas
4 Operator
•
1.5K Posts
0
July 10th, 2008 12:00
Thanks for reaching out to us
Cheers,
Sandip
jimkunysz
259 Posts
0
July 10th, 2008 12:00
I think I might have answererd my own question with an alternate solution.
I can configure the standby dm03 with the correct network cabling and etherchannel configuration and simply failover the prod dm02. then reconfigure the prod dm02.
nandas
4 Operator
•
1.5K Posts
0
July 10th, 2008 12:00
However, few things to consider -
1) I am assuming that currently only cge0 is being used (as the valid connection in the trunk) for all user connections - however, you may confirm that looking at the network switch port activity and throughput. The port to whcih cge 1 is currently connected should not any traffic.
2) when you configure Etherchannel on the switch - it may cause a momentary loss of connectivity for cge0. You may use portfast feature on these ports to minimise the time. UNIX clients will not be affected as NFS will keep on retrying the connection - but windows client may loose their session and have to reconnect to the shares on NAS.
So, better to do it at off-peak hours.
Hope this helps.
Thanks,
Sandip