Subscribe
Accepted Solution

Which API can be used to tell the HA pair nodes are standard or mirrored ones?

 

Re: Which API can be used to tell the HA pair nodes are standard or mirrored ones?

If you are looking for syncmirror status for aggregates, you can take a look at 

https://community.netapp.com/t5/Software-Development-Kit-SDK-and-API-Discussions/7-mode-zapi-call-to-get-a-metrocluster-info/m-p/116754#M1831

for cDOT you can use aggr-get-iter zapi 

 

Re: Which API can be used to tell the HA pair nodes are standard or mirrored ones?

What are you referring to when you say "mirrored" nodes?  Metrocluster?

 

Andrew

If this post resolved your issue, please help others by selecting ACCEPT AS SOLUTION or adding a KUDO.

Re: Which API can be used to tell the HA pair nodes are standard or mirrored ones?

I saw in the documentation that cluster conssists of many HA pairs. Each HA pair has two nodes.

In the standard HA pair, these two nodes both actively provide data service. When one node goes down, the other node will take over the disk from the failed node. the Taked over disk becomes so called partner disk. This two node behaves like active-active failed over pair.

 

Does mirrored HA pair behaves the similar way? the difference is one node has a copy of the othe node's data? I am confused over here. Or Mirrored HA pair works together to provide data mirror for the metro cluster?

 

Thanks!

Re: Which API can be used to tell the HA pair nodes are standard or mirrored ones?

I'm going to assume when you say "mirrored HA pair" you're referring to a metrocluster configuration.  Metrocluster can have multiple fail-over scenarios...

 

Let's assume that we have a simple MCC setup:  Site A has a 2 node HA pair and site B has a 2 node HA pair.  I'll refer to these as "A1" and "A2" for site A, and "B1", "B2" for site B.  I'll also assume that site A is the primary/production site, while site B is the DR site.

 

  • HA pair "A" can have a node fail, resulting in A1 hosting both it's workload and A2's workload (this is standard failover).  The inverse is also true...A2 could take over for A1, resulting in it hosting both workloads
  • Just as above, HA pair "B" can have a node fail.
  • "A" can failover to "B"
  • A1 can failover to A2 (or vice-versa, A2 could failover to A1), then A can failover to B
  • A can failover to B, then B1 could then failover to B2 (or vice-versa).  This would result in a single node now servicing the workload for all four nodes

Definitely check out the documentation if you want more information.

 

If you want to programmatically get information about the node/cluster configuration (and status), use the metrocluster-get and system-get-node-info-iter APIs.

 

Hope that helps.

 

Andrew

If this post resolved your issue, please help others by selecting ACCEPT AS SOLUTION or adding a KUDO.