How does dbsrv12 -im c decide when to take a checkpoint?

[Breck_Carter]

To ask the question another way...

Is it unusual, or expected, for a version 12.0.1.3457 dbsrv12 running with -im c on RedHat ES5.7 to have 0% checkpoint urgency throughout a 16 minute period of heavy activity, with no checkpoints taken during that period?

Yes, it's not a very good sample, especially since it hasn't passed the standard-20-minute-boundary, but... I'm used to the checkpoint urgency increasing over time.

And with -im c checkpoint-only-in-memory-mode it's those checkpoints that give you the only recoverability you have.

---

...tickle edit

[Former Member · Accepted solution]

According to the documentation for -im c, the checkpoint frequency is the same as when running with a log file (even though there is no log file), as described here.

[former_SQLA_member1695117]

With In-Memory CheckPoint Only (-im c), I started the server with -gc option (or without), but does not perform automatic checkpoint. Only when shutdown or "checkpoint" command.

What is missing?

I use sql anywhere 16.0.0.2178