SAP Consultants force to backint backup. So they accept this method, regardless of whether they are virtual or physical. Because this is the certified method.
What happen If HANA runs on VM and customer Veeam license is socket based?
Customer have to buy Instant based licence or socket based Enterprise Plus license is enough for using backint?
This is just a SAP community wiki maintained by God knows who
The official source of this information is SAP Certified Solutions Directory, and as you can see Veeam is listed there with the certification dated back to November 2018.
We are configured at customer HANA Plugin. Jobs are successful but email notifications can not work. Only virtual machine and agents mail are coming. Is Veeam can not send notifications HANA jobs?
I apologize for the confusion, now it's all clear.
By default, we do not send "success" notifications on Veeam Plugin (RMAN, HANA) jobs to not spam your mailbox. This can be changed by modifying the below reg key on the backup server. The reboot/services restart is not required.
PluginEmailNotifications
HKEY_LOCAL_MACHINE\SOFTWARE\Veeam\Veeam Backup and Replication\
Thank you Fedor. I'm change the registry value and everythings fine!
I want to ask SAP HANA Plugin Proxy Usage. I think compression is doing via plugin so Plugin don't use proxy processor.
But i don't know concurrent job behavior.
Customer have Veeam Console VM, and ScaleOutRepository VM. These two VM's also HotAdd proxy. Virtual machine backups & replication jobs also use these proxies. What happen all proxies are busy?
Because SAP HANA Plugin connect at first Veeam Console VM, after that Veeam Console VM transferred data to ScaleOutRepository VM.
Which proxy is always empty for SAP HANA full fast/performance backup?
You are welcome, Mehmet. Glad to see it works for you now.
Regarding HANA Plugin and Backup Proxy: HANA Plugin jobs do not utilize Backup Proxies at all. HANA Plugin contains a data mover (transport agent) package itself that is deployed on HANA DB server and target repository server.
tsightler wrote: ↑Feb 08, 2016 6:51 pm
Yes, it is valid for replication as well. Note that when you fail over the replica VM and the HANA services start they will automatically see the snapshot, perform a recovery and the snapshot will be tagged as "abandoned due to service restart". This is expected behavior and is the same behavior seen during full VM restores as well. No user interaction is required.
I've continued work on these scripts and they've grown significantly more sophisticated. Below I will post the URL to the current versions. This version has a few variables for username, password, SAP path (path to sapservices file) and timeout (default 5 minutes). The scripts now automatically parse the sapservices file to find installed instances, place each in snapshot mode, wait and check the status of the snapshot prior to exiting the pre-freeze script. They also no longer require storing a temporary file on the filesystem. I still plan to add a few more enhancements, such as the ability to use the user secure store instead of a password, and possibly some log file management, although the later can also easily be handled with a simple cron job so that might be overkill.
Please test and let me know how they work. Feedback and bug reports are welcome.
They are still working. As SAP HANA support now multi tenancy with SnapShot processing, you need to loop those for the different databases, but beside this it will just work.
We are working on some enhancements, but overall they are up to date.
As I maintain the script I did some work on supporting multi-tenant but ran into a few complex challenges. The original script was not written with anticipation of multi-tenancy being supported for snapshots and adding it as is turned out to add more complexity to the script that made debugging very difficult. Ideally, to make the code supportable, I need to make a few architectural changes such as turning certain portions of the script into functions. If there is interest to support multi-tenancy and you are willing to test the script please open an issue on the GitHub repo requesting this feature and I will restart work on it.
Maybe we should elaborate first on a more general level. I spoke to several HANA consultants and they tried to convince me that even a crash-consistent HANA image-level backup would be usable right away due to the way HANA deals with its commits.
So when using the Veeam Backint plugin in addition anyways for DB/log backups - would you see the requirement to use these scrips at all?
How do other larger HANA customers handle that?
SAP HANA support only cares about "backups" that are performed and visible in their backup catalog.
You need File/BACKINT backups mandatory (for example to store the automatic log backups).
Then everything else is nice to have from their side. SAP will not support any backup method that is not in the backup catalog visible.
In general Crash consitent situation (for example power loss or restore of a crash consistent snapshot) should not be a problem for HANA istelf. But HANA will start a self healing process that can last multiple days depending on the size of the database and disk speed when you start from a crash consistent state. It is way faster to go the "SAP HANA Snapshot preparation" path where the restored backup/snapshot would just start in near usual start time.