From ea70a3af540b8091884624f080d3c7e6d0665fee Mon Sep 17 00:00:00 2001 From: iblutrifork Date: Tue, 3 Oct 2023 15:08:30 +0200 Subject: [PATCH] Update docs --- charts/flink-job/README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/charts/flink-job/README.md b/charts/flink-job/README.md index afce569b..bff11923 100644 --- a/charts/flink-job/README.md +++ b/charts/flink-job/README.md @@ -154,7 +154,7 @@ Read more about Flink and highly available job-managers [here](https://nightlies | job.topics | list | `[]` | Define which topics this job will consume. Used for data-discovery in Cheetah Backstage. If the `arg` variable is set, adds `-- ` to the arguments passed to the job If `arg` is not set, adds `---kafka-topic ` or `---kafka-topic- ` See [values.yaml](values.yaml) for the format | | job.parallelism | int | `1` | How many jobs to run in parallel, see more here: | | job.state | string | `"running"` | Desired state of the job. Must be either: `running` or `suspended` | -| job.upgradeMode | string | `"savepoint"` | Application upgrade mode. Must be either: stateless, last_state, savepoint `stateless` upgrades is done from an empty state `last-state` does a quick upgrade. Does not require the job to be in a healthy state, as it makes use of the HA metadata `savepoint` makes use of savepoints when upgrading and requires the job to be running. This provides maximal safety Read more here: | +| job.upgradeMode | string | `"savepoint"` | Application upgrade mode. Must be either: stateless, last-state, savepoint `stateless` upgrades is done from an empty state `last-state` does a quick upgrade. Does not require the job to be in a healthy state, as it makes use of the HA metadata `savepoint` makes use of savepoints when upgrading and requires the job to be running. This provides maximal safety Read more here: | | job.savepointTriggerNonce | int | `0` | change this to trigger a savepoint manually, see more here: | | job.initialSavepointPath | string | `""` | change this to force a manual recovery checkpoint, see more here: | | job.allowNonRestoredState | bool | `false` | If this is true, it will ignore the past checkpoints and start anew. Usefull if the job schema has changed. |