We are using multiple instances of Octopus on one server. This because of the fact that our customers have different DTAP streets and projects. Therefore we made the decision to try to use an instance per customer.
Anyway, we noticed that a simple instance when idle eats up 350MB. Is this normal? Seems like an afwul lot to me?
This will limit the number of instances we can run simultaneously on one server.
Thanks for getting in touch!
Firstly, how many instances are you running ? As per our pricing page:
“The EULA allows you to use the same license key on up to three separate Octopus Deploy server instances. This means you can use one for pre-production, one in an isolated production data center, and another for testing at no extra cost. If you need more than three, contact us for discounted licensing.”
Regarding the memory, do your deployments output a lot of log messages ? We’ve just fixed a bug which we released as part of 2.6.1 which means that we should use a lot less memory (we were doing something silly when returning log data).
Hope that helps!
Currently only 2 instances running. The instance I am talking about only has 1 project, 4 environments and is using the built in NuGet feed. And has 8 deployments steps, pretty basic stuff.
I yesterday upgraded to 18.104.22.1686 and the diagnostics page is telling me that this instance (which is idle) is using 370MB, so no deployments or NuGet pushes going on.
What can be expected regarding memory usage of an instance?
Octopus while idle can be anything up to 500MB. It is due to the embedded database. When 3.0 comes around much of this will be transferred to SQL, which could also be on a different server.
Hope that helps!