In various instances, the necessity arises to conduct meticulous maintenance on the SAP HANA database, whether it be for operating system enhancements or database updates, among other considerations. It is imperative that such maintenance activities align seamlessly with your SAP systems, especially when implemented in conjunction with a High Availability/Disaster Recovery (HA/DR) solution running on MS Azure.
Let's contemplate a scenario wherein your operational framework aligns with a three-tier landscape configuration, delineated as follows -
HANA DB Architecture
Site1 - Primary (Running on Azure Region 1 - Zone 1)
Site2 - Secondary (Running on Azure Region 1 - Zone 2)
Site3 - DR(Disaster Recovery running on Azure Region 2)
HANA DB - SAP HANA SPS06 DB including XSA (RHEL 8x)
Below 2 parameters are active in DB -
[inifile_checker]/enable = true
[inifile_checker]/replicate = true
Replication Scenario (configured as per requirement)
Site1(Primary) >>> SYNC Replication >>> Site2(Secondary)
Site2(Secondary) >>> ASYNC Replication >>> Site3(Disaster Recovery)
XSA Configuration -
Default Domain - Hostname<Site1>
API_URL - Hostname<Site1>:port
You have completed HSR setup as per below URL's and it's successfully tested as well -
https://learn.microsoft.com/en-us/azure/sap/workloads/sap-hana-high-availability-rhel
https://learn.microsoft.com/en-us/azure/sap/workloads/disaster-recovery-sap-hana
During approved maintenance, you need to handover the VM's to OS team for patching and reboot.
Assumptions and Prerequisites for SAP HANA Database Maintenance -
In order to mitigate the risk of a Single Point of Failure, it is assumed that at all times, two database (DB) servers should maintain an active replication status. This strategic redundancy ensures a robust and fault-tolerant system architecture.
Before embarking on the maintenance journey, certain prerequisites must be diligently addressed:
1. Replication Vigilance:
- It is essential to consistently validate the replication status within HANA Studio and at the operating system level. This verification process ensures that the replication mechanism is not only active but also performing optimally.
2. WMM Configuration in Solution Manager:
- To preemptively counter potential alerts, it is imperative to configure Work Mode Management (WMM) in the Solution Manager. This proactive measure aids in maintaining a harmonious operational environment and minimizes disruptions during the maintenance phase.
3. Mitigating External Monitoring Alerts:
- To streamline the maintenance process further, consider suppressing alerts from third-party monitoring tools. This precautionary step prevents unnecessary interruptions, allowing for a focused and unencumbered maintenance operation.
HANA DB DR(Site3) Server -
hdbnsutil -sr_register --name=Site3 --remoteHost=<Site1 host> --remoteInstance=<Site1 instance nu> --replicationMode=async --operationMode=logreplay
#[ha_dr_provider_SAPHanaSR]
#path = /hana/shared/myHooks
#execution_order = 1
#provider = SAPHanaSR
sapcontrol -nr <instance num> -function GetProcessList
New Replication Scenario -
Site1 >>> SYNC Replication >>> Site2
Site1 >>> ASYNC Replication >>> Site3
HANA Secondary (Site2) Server -
pcs property set maintenance-mode=true
sapcontrol -nr <instance num> -function GetProcessList
pcs property set maintenance-mode=false
No Change in Replication Scenario -
Site1 >>> SYNC Replication >>> Site2
Site1 >>> ASYNC Replication >>> Site3
Now we are moving to Primary DB restart hence it's requires to change the replication setup of DR system from Site1 to Site2 to keep 2 systems replication active.
This step is optional, as per requirement -
hdbnsutil -sr_register --name=Site3 --remoteHost=<Site2 host> --remoteInstance=<Site2 instance nu> --replicationMode=async --operationMode=logreplay
#[ha_dr_provider_SAPHanaSR]
#path = /hana/shared/myHooks
#execution_order = 1
#provider = SAPHanaSR
sapcontrol -nr <instance num> -function GetProcessList
New Replication Scenario -
Site1 >>> SYNC Replication >>> Site2
Site2 >>> ASYNC Replication >>> Site3
HANA Primary (Site1) Server -
pcs resource move <resource_name> <node_name>
pcs cluster status (if secondary resource is in stopped status, keep in stopped status)
pcs property set maintenance-mode=true
hdbnsutil -sr_register --name=Site1 --remoteHost=<Site2 host> --remoteInstance=<Site2 instance nu> --replicationMode=sync --operationMode=logreplay
sapcontrol -nr <instance num> -function GetProcessList
pcs property set maintenance-mode=false
New Replication Scenario -
Primary Site2 >>> SYNC Replication >>> Site1 Secondary
Primary Site2 >>> ASYNC Replication >>> Site3 DR
You can keep the Primary Site2 and change the DR replication from Site2 to Site1 as per above given DR replication steps or else changing the Primary Site from Site2 to Site1 will follow all the switchover steps.
Troubleshooting -
1. XS controller service is not starting up post primary switchover due to old host default domain and api_url (as per configuration)
To fix this situation, follow the below steps - (this will not stop the INDEXSERVER service on primary system)
XSA disable
hdbnsutil -sr_unregister
Disable system replication on the primary system
hdbnsutil -sr_disable
Change the Default Domain and API_URL to local hostname
XSA enable
hdbnsutil -sr_enable --name=Site2
2. Error while registering DR from Standby/Primary Server
Error -
INFO: no copy check for systemPKI SSFS files as primary does not support this (older version than secondary) or primary has deactivated it explicitly
got request for system replication site registration, but system replication is not enabled.
To fix this situation, follow the below steps -
hdbnsutil -sr_enable --name=Site2 or Site1
References -
2964676 - How to disable XSA (SAP HANA XS Advanced)
https://help.sap.com/docs/SAP_HANA_PLATFORM/2c1988d620e04368aa4103bf26f17727/73596bf87326455e8f2c10b...
https://blogs.sap.com/2020/05/27/how-to-change-sap-hana-cockpit-url-from-physical-hostname-to-a-dns-...
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
10 | |
6 | |
5 | |
5 | |
4 | |
4 | |
4 | |
4 | |
3 | |
3 |