chore: improve documentation on post/pre-update commands

This commit is contained in:
Alexandre Menif 2018-10-25 20:52:36 +02:00
parent 3df7bbfa32
commit 0870d7b1c1

View file

@ -277,12 +277,18 @@ docker run -d \
``` ```
## Executing commands before and after updating ## Executing commands before and after updating
For every container that could be updated by watchtower, it is possible to It is possible to execute a *pre-update* command and a *post-update* command
specify a command that will be executed before stopping the container (a inside every container updated by watchtower. The *pre-update* command is
`pre-update` command), and a command that will be executed after restarting the executed before stopping the container, and the *post-update* command is
container (a `post-update` command). These commands are specified using the executed after restarting the container.
*com.centurylinklabs.watchtower.pre-update-command* and the
*com.centurylinklabs.watchtower.post-update-command* labels. Both commands are shell commands executed with `sh`, and therefore require the
container to provide the `sh` executable.
The commands are specified using docker container labels, with
*com.centurylinklabs.watchtower.pre-update-command* for the *pre-update*
command and *com.centurylinklabs.watchtower.post-update-command* for the
*post-update* command.
These labels can be declared as instructions in a Dockerfile: These labels can be declared as instructions in a Dockerfile:
@ -300,3 +306,6 @@ docker run -d \
someimage someimage
``` ```
The failure of a command to execute, identified by an exit code different than
0, should not prevent watchtower from updating the container. Only an error
log statement containing the exit code will be reported.