SVM Peer relationship

[SrinathATT]

SVM(vserver1) is in peered with vserver2 and and vol1 is in snapmirrored with vol1_dr,

If source vol(vol1) is unavailable vol1_dr can serve the data to the source SVM as Netapp says.

Can u pls post the cmds and process how do we do it.

what is the role of Cluster peer, SVM peer to setup snapmirror when compared to 7 mode?

--Srinath

[mikkisse]

If i remember correctly, you must:

1. Quiesce transfers and break snapmirror relatonships from destination side, after that your destination volume will be rw.

2. You must reconnect your clients to destination svm to serve data. 

3. After recovering source vol, you can reverse resync data from destination volume to source.

4. Break relationships again, but now from your source after syncronization and reconnect clients to source svm with restored vol.

5. Create resync again from source volume to destination volume. 

All operations you can perform from web client. 

Sorry, if there something wrong. I haven't practice with snapmirror 😞 

[asulliva]

Clustered Data ONTAP is a multi-tenant system, the cluster and SVM peering system supports this.  The cluster administrator configures peering relationships at the cluster level, followed by SVM administrators configuring them at the SVM layer.  With this division of labor, a tenant who has multiple SVMs on multiple clusters can manage their own peering and configuring the SnapMirrors, while the cluster admin is only responsible for ensuring that the underlying infrastructure (intercluster links, cluster peers, etc.) is configured.

To change a destination from read-only to read-write you'll need to do the following from the cluster shell:

# from the destination, stop future transfers of the snapmirror
snapmirror quiesce -destination-path vserver:volume

# break the relationship, making the destination read-write
snapmirror break -destination-path vserver:volume

# junction the volume
volume mount -vserver vserver -volume volume -junction-path "/volume"

# at this point the clients would need to mount the volume from the 
# new source, the destination SVM NFS LIF(s)

Andrew

[jorgeedugona]

When a source volume becomes unavailable for read/write access, you can perform a volume-level disaster recovery failover and resynchronization from OnCommand System Manager. This functionality is available from the Protection page of the destination Vserver. For further info see the below links:

https://library.netapp.com/ecm/ecm_doxwnload_file/ECMP1287556

https://library.netapp.com/ecmdocs/ECMP1242991/html/GUID-8B786DDE-5230-4C69-8DC4-575FF85615D6.html