I note that when autocheckpointing in turned on, checkpoints will be run W

RE: I note that when autocheckpointing in turned on, checkpoints will be run W

Eliot Gerson — Wed, 07 Mar 2018 18:46:38 GMT

The documentation here docs.appian.com/.../Configuring_Application_Checkpointing.html includes the current (since 17.3) guidance for checkpointing and instructions for configuring it.

RE: I note that when autocheckpointing in turned on, checkpoints will be run W

nathan.schmitz — Wed, 07 Mar 2018 18:38:46 GMT

In the server/_conf directory update the gateway-config.xml file to set the auto-database-save="0" to disable auto check pointing.

RE: I note that when autocheckpointing in turned on, checkpoints will be run W

joanneh — Wed, 07 Mar 2018 08:00:19 GMT

Hi Mike Cichy

Do you know how to disable auto checkpointing?

RE: I note that when autocheckpointing in turned on, checkpoints will be run
W

Mike Cichy — Thu, 23 Aug 2012 22:03:33 GMT

1) You can easily tell by the kdb last modified timestamp. Like I said, these times will vary on a number of factors.
2) You will need to inspect the db logs, and calculate the time between initiation of the save to the kdb switch. Normally the save is limited by I/O because of writing the sizeable files to disk but should take less than a few seconds on DEV. You can open an Appian Support case to help you identify the start and end times.

RE: I note that when autocheckpointing in turned on, checkpoints will be run
W

michaelr236 — Thu, 23 Aug 2012 19:55:08 GMT

Thanks Mike. In Dev, where autocheckpointing is enabled, how can we find 1) how frequently checkpoints occur, and 2) how long a checkpoint takes?

RE: I note that when autocheckpointing in turned on, checkpoints will be run
W

Mike Cichy — Wed, 30 May 2012 07:31:16 GMT

We recommend that auto checkpoint be disabled for production systems even for multiple gateway installations. With auto checkpointing, you are able to set the thresholds but you are unable to control when the thresholds are reached.
As to how long a checkpoint runs, depends on mutliple factors: size of the kdb, disk I/O, etc. but anything above 10 seconds is logged as a warning in the logs.