on 06-10-2021 10:55 AM
Hello Community,
we want to transfer a HANA Database from one data center to another one minimizing downtime. Due to a large database and limited network ressorces we want to proceed as follows:
Preparation:
Full backup to file
Transfer via HDD
Recover to target
n-times until reaching planned downtime:
{Inkremental Backup, Transfer via Network, Recover inkremental}
Downtime
Stop Source
Inkremental Backup, Transfer via Network, Recover inkremental
We are using this procedure sucessfully with MaxDB and Oracle but I cannot find the way with HANA.
I have found a threat of 2016 with the information, that backup catalog is required to start the recovery. But at the time of the recovery of the full backup, the catalog can not have the information of the last inkremental backup.
When I use HANA Studio the database will be started after recovery of a full backup without catalog. I can not start the recovery of the inkremental recovery after that.
I have tried to find a recover command with sql, but did not find a usable.
I need this procedure urgently to reduce downtime from around 24 hours (backup to disc, transfer, recovery in downtime) to nearly one hour by described way.
Can anyone give me a hint, where to find the right tool or command?
Best Regards
Andreas
Hi Andreas,
For every data, delta (incremental & differential) and log backup , the nameserver of the database writes backup catalog (log_backup_0_0_0_0.<BackupID> at location : global.ini/[persistence]/basepath_catalogbackup). This is true even in situations when log mode overwrite is set.
The purpose of incremental and log backups is to recover the database to most possible recent point. Thus, the delta and log backups support the full backup to achieve the recent state.
HANA gives only two option
- Recover the database with full backup : without catalog.
- Recover the database with full backup + delta backup (if available) + log backup : With source backup catalog (and delta/log files).
In your case, you may proceed with recovery using backup catalog only. Ensure that you don't miss any backup files or required catalog file.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hello,
the fastes way is then:
I think for comfort SAP should develop the possibility to recover the database the same way she did with MaxDB This would reduce downtime of system copies and transfers alot. I do not get why she did not.
A last question, can I use replication of hana without direct network connection? Perhaps with full backup and logshipping via file transfer?
The point is:
And I have a productive system with material managent and sales working 7x24. So I have to minimize downtime.
Best Regards
Andreas
Hi Andreas,
Instead of recovering multiple incremental backups until your planned schedule, you can set-up HANA system replication(HSR) between the two systems to perform the copy and take-over at your planned schedule. The continuous log replay will ensure that the target is ready at any time.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
95 | |
11 | |
10 | |
9 | |
9 | |
7 | |
6 | |
5 | |
5 | |
4 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.