Retrying and limiting messages length when sending to teams

[blalop]

Description of the Change

In the current state of prom2teams, there is nothing it can do to prevent missing Prometheus alerts due to several reasons, such as MS Teams being not available or prom2teams trying to send a payload that exceeds the limit MS Teams has.

Benefits

No alerts will be lost due to:

• Teams being down
• Sending a payload that is too long

Possible Drawbacks

None

Applicable Issues

#223

[vicsufer]

Question time, how is this interacting when using multiple connectors?

• The same configuration is applied for all of them? (I think so)
• If a request fails into one connector, the retry is only applied to this connector or will it trigger an unnecesary retry to the others?

[blalop]

Hi @vicsufer:

• You're right, this PR does not provide per-connector configuration.
• As the retry is managed at the teams client level - it has no knowledge of connectors. This means that it will only retry the failed connections.