Runbooks can be run via a scheduled Trigger. This is a very useful feature. But it would be nice if the Trigger had a feature to help when the triggered RunBook Run fails.

I would like to propose that another field be added to the Trigger creation page. It could be called "Error Cron Expression".

When the RunBook Run by the Trigger fails, the Trigger would switch its Cron Expression to use the "Error Cron Expression". Once the Trigger starts succeeding, then it would switch back to the normal Cron Expression.

The idea is that, when a Trigger is failing, it can be useful to run it at a different frequency. Sometimes you want to back off to not overwhelm a system with failed calls. If the Trigger is a "once a day" kind of trigger, you may want to run the Trigger every hour till is starts succeeding (since a failed "once a day" trigger means it will not try again till the next day).

Either use case would be fulfilled by an "Error Cron Expression" feature.

It is great that you can now centrally manage your certificates from within Octopus Deploy. I am using this for managing the bound https cert in IIS sites and it works great.

One drawback we've found is that if a certificate is replaced, you have to re-deploy the project again to apply the new certificate. Our deployment process is quite lengthy and involved, turning services on and off causing at around 10-15mins of downtime of our system.

Ideally, we would be able to push a new certificates to any tentacle that has the old certificate, without triggering a deployment of all the services. Bonus points if this also updated the new certificate thumbprint in IIS as well.

It might be expected that you must re-deploy to install the new certificate, I just personally find that it causes significantly more down-time that it could be, if only the certificate was pushed.