Skip to main content

Community

Connector Improvement: HANA Archive Only from Different Node

Answered

Please sign in to leave a comment.

Comments

3 comments

    It should be possible to configure an external location for the logs,

    You can do that by setting the  Archive_Log_Path and Archive_Log_Format (https://fivetran.com/docs/hvr6/property-reference/location-properties#locationproperties) to point to archives.

    Edwin, What i'm referring to is not the Archive_Log_path.  With HANA, if you're using backint to capture the logs, HVR queries the backup catalog.  And it assumes that the backup catalog is available on the location that is configured for the channel.  Due to resource constraints, we have to run the HVR agent on a backup node, which does not have the backup catalog available to it, because it's the active node that manages the backup catalog. So having a setting available to specify the location of the backup catalog, allows us to simplify the HVR channel.  Currently we have two source locations configured.  One is for refreshes only and the other is only used during the activation of replication (backup catalog).  And we already have it set to use the backup log only using backint.  So it uses the node local to the agent to "restore" the backup log and read from it.

    ok, I get it now. Alright, I will feed this into the Feature request.