Set up High Availability between 2 HANA DB Sites

In the last 2 posts we saw how to install HANA DB on 2 nodes and how to setup site replication between those 2 nodes here and here

In this post we will see how to set up HA between both sites.

Please Note: This guide is prepared with Reference from the official documentation here

To setup HA please have SBD configured on both nodes.

For this setup I have used multipath for consistency.

Configure Multipath

First make sure you have an Iscsi server which exposes a device of 2 GB as Iscsi target.

In both nodes make sure Iscsi client is configured and connected.

To setup iSCSI in SLES for SAP refer the official guide here

Once configured you should be able to see the device in the device list

fdisk -l

Next install multipath if not already installed

zypper install multipath*

Start and enable multipathd.service

systemctl enable multipathd.service
systemctl start multipathd.service

Start and enable chronyd.service

systemctl enable chronyd.service
systemctl start chronyd.service

Generate the multipath configuration file using the command

multipath -T >/etc/multipath.conf

This command should automatically configure multipath for the sbd device, however alias won’t be added.
Edit the multipath.conf file and add the alias for the device.

Reload Multipath Service

service multipathd reload

New device should be available in fdisk -l

This is the device we will use in HA configuration.

Repeat the steps on Node 2

HA installation

Initiate HA installation on Node 1

ha-cluster-init -U -s /dev/mapper/mpath-sbd

Answer “No” for VIP and QDevice.

Commands to check if SBD has been configured correctly

egrep -v "(^#|^$)" /etc/sysconfig/sbd
sbd -d /dev/mapper/mpath-sbd dump
sbd -d /dev/mapper/mpath-sbd list

On second Node to join this cluster

ha-cluster-join -c 192.168.211.42

To check the cluster: (On both nodes)

systemctl status pacemaker
systemctl status sbd
crm cluster start
crm status

HANA Cluster Configuration

Now that the initial cluster is setup lets configure SAP HANA multistate using separate configuration files and loading it into cluster.

The syntax would be as below

vi crm-fileXX
crm configure load update crm-fileXX

Cluster bootstrap

vi crm-bs.txt
###
# enter the following to crm-bs.txt
property cib-bootstrap-options: \
    stonith-enabled="true" \
    stonith-action="reboot" \
    stonith-timeout="150" \
    priority-fencing-delay="30"
rsc_defaults rsc-options: \
    resource-stickiness="1000" \
    migration-threshold="5000"
op_defaults op-options: \
    timeout="600" \
    record-pending=true

Load config to cluster

crm configure load update crm-bs.txt

STONITH Device

vi crm-sbd.txt
###
# enter the following to crm-sbd.txt
primitive stonith-sbd stonith:external/sbd \
    params pcmk_delay_max="15"

Load Config to Cluster

crm configure load update crm-sbd.txt

SAP HANA Topology

vi crm-saphanatop.txt
###
# enter the following to crm-saphanatop.txt
primitive rsc_SAPHanaTop_HA1_HDB10 ocf:suse:SAPHanaTopology \
    op monitor interval="10" timeout="600" \
    op start interval="0" timeout="600" \
    op stop interval="0" timeout="300" \
    params SID="HA1" InstanceNumber="10"
clone cln_SAPHanaTop_HA1_HDB10 rsc_SAPHanaTop_HA1_HDB10 \
    meta clone-node-max="1" interleave="true"

Load config to cluster

crm configure load update crm-saphanatop.txt

SAP Hana

vi crm-saphana.txt
# enter the following to crm-saphana.txt
primitive rsc_SAPHana_HA1_HDB10 ocf:suse:SAPHana \
    op start interval="0" timeout="3600" \
    op stop interval="0" timeout="3600" \
    op promote interval="0" timeout="3600" \
    op monitor interval="60" role="Master" timeout="700" \
    op monitor interval="61" role="Slave" timeout="700" \
    params SID="HA1" InstanceNumber="10" PREFER_SITE_TAKEOVER="true" \
        DUPLICATE_PRIMARY_TIMEOUT="7200" AUTOMATED_REGISTER="false" \
    meta priority="100"
ms msl_SAPHana_HA1_HDB10 rsc_SAPHana_HA1_HDB10 \
    meta clone-max="2" clone-node-max="1" interleave="true" maintenance=true

Load config to cluster

crm configure load update crm-saphana.txt

Virtual IP Address

# vi crm-vip.txt
# enter the following to crm-vip.txt
primitive rsc_ip_HA1_HDB10 ocf:heartbeat:IPaddr2 \
    op monitor interval="10s" timeout="20s" \
    params ip="192.168.1.20"

Load config to cluster

crm configure load update crm-vip.txt

Activating multi-state resource for cluster operation

crm resource refresh msl_SAPHana_HA1_HDB10
crm resource maintenance msl_SAPHana_HA1_HDB10 off

That’s it you successfully configured HANA with replication and HA.

If you go to your HAWK portal, you should see as below:

Testing

Test 1:

Stop Hana on node 1

HDB stop

Automatic take over should happen and node 2 should be promoted.

Refresh cluster resource and check status

crm resource refresh rsc_SAPHana_HA1_HDB10 hana-n1
crm status

Node 1 should start and stay as secondary

To Recover

hdbnsutil -sr_register --name=WDF --remoteHost=hana-n2 --remoteInstance=10 --replicationMode=sync --operationMode=logreplay

Test 2

Stop Hana on node 2

HDB stop

Automatic take over should happen and node 1 should be promoted

To Recover on Node 2

hdbnsutil -sr_register --name=ROT --remoteHost=hana-n1 --remoteInstance=10 --replicationMode=sync --operationMode=logreplay

Node 2 should start and stay as secondary

That is it, we have successfully setup High Availability for a 2-Node HANA DB Cluster with Site Replication.