Problem Note 65233: Failed Postgres nodes re-entering the cluster might take a long time to restore the data
If a failover of the SAS® Infrastructure Data Server (PostgreSQL database) in SAS® Viya® 3.5 occurs, you must fix the issue with the node that went down before having it re-enter the cluster. After you fix the node and the node attempts to re-enter the cluster, the data could take a long time to restore.
Click the Hot Fix tab in this note for a link to instructions about accessing and applying the software update.
When you apply this hot fix, any SAS Infrastructure Data Server using PostgreSQL 11.5 or later uses the pg_rewind utility to synchronize the cluster with another copy. Using the PostgreSQL pg_rewind utility instead of the pg_basebackup utility that is used for PostgreSQL 9.4 decreases the time needed to restore the cluster.
You must set the following parameters to on in the postgresql.conf file:
wal_log_hints = on
full_page_writes = on
Note: By default, they are set to on.
Operating System and Release Information
SAS System | SAS Viya | Linux for x64 | 3.5 | | Viya | |
*
For software releases that are not yet generally available, the Fixed
Release is the software release in which the problem is planned to be
fixed.
Type: | Problem Note |
Priority: | high |
Date Modified: | 2019-12-18 09:40:34 |
Date Created: | 2019-12-10 12:57:27 |