HANA Near Zero Downtime Maintenance Activity

In this article, I am going to explain how to perform near zero downtime HANA maintenance activities in 4 tier HANA Multi Target setup.

Now days for customers main challenge is near zero downtime and with this objective – I performed HANA Node Operating System Maintenance activity in 4-tier HANA setup near zero downtime.

There was no impact to Business Operation during this major maintenance activity.

In Current setup, we have following replication setup in 4 nodes.  

Step1: Check existing Replication setup

Replication should be in Active state for 4 nodes.

Step 2:

Placing a Pacemaker cluster into maintenance mode or freezing the cluster on HANA Database – Node A & Node C

pcs property set maintenance-mode=true

Step 3:  Stop HANA database on HANA Node D, Node C and Node B and unregister HANA Database

hdbnsutil -sr_unregister (in each HANA Node – Node D, Node C and Node B)

Step 4: Disable Replication on Node A:

hdbnsutil -sr_disable

Step 5: Handover To Operating System Team for planned maintenance activities – HANA Node B , Node C and Node D

Step 6: Once maintenance activities are done for HANA – Node B, Node C and Node D ,

Register Node B to Node A:

hdbnsutil -sr_register –name=<Node B> –remoteHost=<Hostname of HANA Node A > –remoteInstance=<Instance No> –replicationMode=syncmem –operationMode=logreplay

 

Step 7: Once Replication is active – Move HANA resources from HANA Node A to Node B.

pcs resource move <Hana Resource Group> <Hostname Node B>

Step 8: Handover To Operating System Team for planned maintenance activities -Node A.

Step 9: Once HANA Node A is back from maintenance activities – Register Node A , Node C and Node D to Primary HANA Node B using following command.

Node A :

hdbnsutil -sr_register –name=<Node A> –remoteHost=<Hostname of HANA Node B > –remoteInstance=<Instance No> –replicationMode=syncmem –operationMode=logreplay

Node C:

hdbnsutil -sr_register –name=<Node C> –remoteHost=<Hostname of HANA Node B > –remoteInstance=<Instance No> –replicationMode=async –operationMode=logreplay

Node D:

hdbnsutil -sr_register –name=<Node D> –remoteHost=<Hostname of HANA Node C > –remoteInstance=<Instance No> –replicationMode=syncmem –operationMode=logreplay

Step 10: Check the replication status – It should be Active for all 4 HANA Nodes

Step 11: Placing a Pacemaker cluster into maintenance mode – Node B

pcs property set maintenance-mode=false