Speed Up Checkpoint Running Time On Solaris For Huge In-Memory Database Caches
(Doc ID 433895.1)
Last updated on APRIL 01, 2018
Applies to:Berkeley DB - Version 4.4.16 to 4.4.20
Oracle Solaris on x86 (32-bit)
Oracle Solaris on SPARC (64-bit)
Oracle Solaris on x86-64 (64-bit)
Oracle Solaris on SPARC (32-bit)
Solaris Operating System (x86)Solaris Operating System (SPARC 64-bit)Solaris Operating System (SPARC 32-bit)
This document is aimed at helping a Berkeley DB 4.4 user speed up the checkpointing completion time when the application presents the following characteristics:
Huge environment database cache configured, more than 4GB in size;
Cache resides on a tmpfs filesystem, where the filesystem is usually a ZFS pool;
Very few dirty pages accumulate in the cache between consequent checkpoints, checkpoints being run very frequently;
The environment is configured with the following flags: DB_CREATE | DB_INIT_MPOOL | DB_THREAD | DB_INIT_TXN | DB_INIT_LOG | DB_INIT_LOCK, and the DB_TXN_NOSYNC flag is set for the environment;
The checkpoint runs in a dedicated thread, and executes every time a timer expires.
To view full details, sign in with your My Oracle Support account.
Don't have a My Oracle Support account? Click to get started!