-
Notifications
You must be signed in to change notification settings - Fork 5.6k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Feature Request - Specific Kafka Offset in telegraf Kafka Consumer #2265
Comments
this could also be solved by solving #2240? |
Hi,
|
As the previous comment pointed out, this is a limitation from the package the plugin is using. Looking at the latest version it doesn't seem this has changed. Closing because there hasn't been any activity in a long time, but if there is someone still interested in feature please comment and we can see what alternatives there might be. Might require an upstream change. |
Feature Request
Proposal:
Currently the Kafka consumer plugin allows you to specify Oldest or Newest as offset values.
It would be very useful to be abler to specify a starting offset number.
Current behavior:
Oldest or Newest specified only work as expected.
Desired behavior:
To specify specify start offset
Use case: [Why is this important (helps with prioritizing requests)]
Kafka is used as a persistent message store to which we send 10's of millions of messages a day. We use telegraf as a consumer to read messages and write them to a number of targets (one being influx).
If any target is unavailable (crashed / maintenance / power failure ) telegraf will drop messages once its buffer is full (which takes a few seconds), and future messages are read from kafka but dropped.
One workaround is to have an additional 'recovery' telegraf agent that re-processes all the messages from kafka. This is started once the target systems that have been unavailable have been brought back on line and can start to accept messages again. Once the backlog has been processed this instance of telegraf is then stopped until it happens again. Each time this is used the recovery process will start from the last offset recorded from the previous use which can be 100's of millions of messages ago or many days / weeks in the past.
It would be useful to be able to specify a starting offset so I only have to re-process messages from the last known good time / offset and not reprocess messages that are known to have been successfully sent to the target system.
The text was updated successfully, but these errors were encountered: