SAP HANA HSR Setup

hana_hsr_setup.sh

DB name: 
	ABC
DB Instance No: 
	01
Site 1 hosts:
	site1node1abc.somedomain.com
	site1node2abc.somedomain.com
Site 2 hosts:
	site2node1abc.somedomain.com
	site2node2abc.somedomain.com


Add below entries in both primary and secondary /etc/hosts file
	#HSR Primary	
	10.200.10.2 hana01-01 site1node1abc.somedomain.com
	10.200.10.3 hana01-02 site1node2abc.somedomain.com
	#HSR Secondary
	10.201.20.2 hana01-01 site2node1abc.somedomain.com
	10.201.20.3 hana01-02 site2node2abc.somedomain.com

Get the IP address of the database host by looking at any one info below 
	(i) nslookup hostname
	(ii) hostname -i 
	(iii) ping hostname 
	(iv) ifconfig -a 
	(v) ip addr -4 a s 

Add respective IP address in /etc/hosts file. We need to add both primary and secondary DB hosts IP mappings in /etc/hosts of both sites.


From Primary, trigger local full database backups for all databases within MDC. This is required for initial data replication to secondary.
You might have minimum 2 DBs (SYSTEMDB, and default Tenant, in this case ABC, if you haven't removed the tenant).
	Please refer: 
	link.medium.com/68Z9iAcFPeb for backups

Primary 
	hdbnsutil -sr_enable --name=SITE1_PRIM_ABC
	# status message 'done' if everything alright

Make sure the Secondary DB cluster is down, if not
	sapcontrol -nr 1 -function StopSystem 

Register secondary using:
	hdbnsutil -sr_register --remoteHost=site1node1abc --remoteInstance=01 --replicationMode=async --operationMode=logreplay --name=SITE2_SEC_ABC
	# status message 'done' if everything alright

Start secondary HANA:
	sapcontrol -nr 01 -function StartSystem 

Once the system started, from Primary, issue below commands to check the status of sync:
	sql> select * from m_service_replication; (check the column REPLICATION_STATUS)
	(or) 
	cmd# cdpy; python systemReplicationStatus.py  (to be run as sidadm OS user, in this case abcadm user)
	(or)
	cmd# hdbcons -e hdbindexserver "replication info" (to be run as sidadm OS user, in this case abcadm user)
	(or)
	cmd# hdbcons -e hdbindexserver "replication info" | egrep 'sentLogPosTimestamp|shippedLogPosTimestamp' (to be run as sidadm OS user, in this case abcadm user)
--------------------------------------


#pitfalls
Hardly you would get any issue while setup, but there are chances. Here is what to look out and remediation

	unable to contact primary site host hana01-01:40102. internal error,location=hana01-01:40102. Trying old-style port (port offset +100)...hana01-01:40102
	(or)
	unable to contact primary site host 10.200.10.2 (hana01-01):40102. internal error,location=10.200.10.2:40102. Trying old-style port (port offset +100)...10.237.58.19 (hana01-01):40102
	==> make sure we have correct hostname/ip mappings setup in DB global.ini file 
		[system_replication_hostname_resolution]
		10.200.10.2 = site1node1abc
		10.200.10.3 = site1node2abc
		10.201.20.2 = site2node1abc
		10.201.20.3 = site2node2abc

	primary and secondary has different systemPKI SSFS data and key or unable to check;Secondary system PKI SSFS-data-file checksum is different from primary!
	==> Starting with HANA 2.0, we need to copy PKI datafile and key (sample below) from Primary to Secondary before we register the secondary. We can SCP/winSCP/SFTP the files for this task. Keep it in same location as primary.
		/usr/sap/ABC/SYS/global/security/rsecssfs/data/SSFS_QAK.DAT
		/usr/sap/ABC/SYS/global/security/rsecssfs/key/SSFS_QAK.KEY

	unable to contact primary site; to 10.200.10.2 (hana01-01):40102; original error: timeout occured,location=10.200.10.2:40102. Was MTU size set to 1500? (Please refer to SAP Note 2142892);
	==> Check the MTU value setup for the NIC interface. It should be 1500 as recommnded for HANA HSR setup. We can keep 1500 for HSR communication interfaces (NICs communicating between primary and secondary).
		ifconfig -a # see the mtu value right next to the interface name
		ifconfig <interface> mtu 1500  # change to 1500, sudo priv necessary. It's online.


Relevant Notes (Incase of any issues, or for your detailed understanding):
	https://launchpad.support.sap.com/#/notes/2772007 - Error : system replication port could not be opened;port=40002
	https://launchpad.support.sap.com/#/notes/2518396 - SAP HANA system replication - registration of secondary site fails
	https://launchpad.support.sap.com/#/notes/2183363 - Configuration of SAP HANA internal network
	https://launchpad.support.sap.com/#/notes/2748924 - Enabling HANA System replication returns error after Hardware change on secondary site
	https://launchpad.support.sap.com/#/notes/2369981 - Required configuration steps for authentication with HANA System Replication
	https://launchpad.support.sap.com/#/notes/2142892 - SAP HANA System Replication sr_register hangs at "collecting information"
	https://launchpad.support.sap.com/#/notes/2081065 - Troubleshooting SAP HANA Network
	https://launchpad.support.sap.com/#/notes/2222200 - FAQ: SAP HANA Network
	https://launchpad.support.sap.com/#/notes/2092196 - How-To: Terminating Sessions in SAP HANA  (for MTU related findings)
	https://launchpad.support.sap.com/#/notes/2166157 - Error: 'failed to open channel ! reason: connection refused' when setting up Disaster Recovery
	https://help.sap.com/viewer/bed8c14f9f024763b0777aa72b5436f6/2.0.02/en-US/0890ba10276e4b0a825640b197be7854.html - Setup and Initial Configuration Problems

HDB Version takes longer time in reporting back the status of systemreplication status

HDB version returns only HANA version details. For replication details, you can issue
python systemreplicationstatus.py
from python directory (Alias to the directory : cdpy)