Redis input plugins label Mismatch about telegraf

[powersj]

Hi,

In this part of the code, the replication_role label is being overwritten as slave. By commenting out this section, the replication_role label correctly reflects master as expected. Here is an example showing the correct behavior after the modification:

The function that you are pointing to parses lines that start with slave (e.g. ^slave\d+). That means from the example output you provided it is parsing the following data:

slave0:ip=127.0.0.1,port=6384,state=online,offset=2870,lag=1
slave1:ip=127.0.0.1,port=6380,state=online,offset=2870,lag=1
slave2:ip=127.0.0.1,port=6381,state=online,offset=2870,lag=1
slave3:ip=127.0.0.1,port=6383,state=online,offset=2870,lag=1
slave4:ip=127.0.0.1,port=6382,state=online,offset=2870,lag=1

I think, Labeling the replication_role as "slave" for metrics collected from the master server might create some confusion.

I can see that, however I think the intention was to clarify that these metrics are representing a slave not a master, not that the data was collected from a master.

I do not see us suddenly changing the value and meaning of a tag that appears to have been there for 5 years without some major churn to existing users.

Thoughts?

[kell0gg]

Thanks for the your comment, I've given it some more thought as you said.

I thought replication_role was not used in other metrics to indicate that the metric is related to a master or slave.

For example, even if there is information related to the master in a node that is a slave, such as the following:

redis-cli -p 6380 info Replication
# Replication
role:slave
master_host:127.0.0.1
master_port:6379
master_link_status:up
master_last_io_seconds_ago:0
...

the replication_role tag is not changed when this metric is collected.

I may also have thought that because I typically use that tag to distinguish between master and replica nodes.

I do not see us suddenly changing the value and meaning of a tag that appears to have been there for 5 years without some major churn to existing users.

I completely agree with your opinion. I think I was too narrow-minded, focusing only on my use case.

I think this issue can be closed. (I'm not sure I understand the issue handling process correctly, so I won't close it for now).
thanks.

[powersj]

Thanks for the response! I will go ahead and close this then.