Backup/restore procedures with ravendb

I want to migrate octopus to another server what is the backup/restore procedures?

I copied c:\Octopus\Data to the new box and the site seems to be up … (have to wait for dns entries) but wanted confirmation that this would work or recommendations.

I notice the PackageCache is several gigs in size… anyway to limit this?


Hi Leblanc,

That works. Sometimes you may have OS compatability issues, in which case you can take one of the backups that Octopus automatically creates, and import it using Raven.Smuggler.exe to import it into a fresh installation. I’ll wrap that tool into a nice Octopus UI shortly.


can you shed some light on C:\Octopus\Applications.Tentacle\Packages - caused me to run out of space on one of my staging environments.

What logic am i affecting by deleting these cached packages? (added logic to postdeploy.ps1 to delete packages older than 10 days )

As you say, that folder is just a cache for downloaded packages - you can delete them with no problems. If they aren’t found they will be re-downloaded.


Raven (6072) C:\Octopus\Data\Data096bbbb3-3a68-4f59-be44-ad37ed424de1: Unable to read the header of logfile C:\Octopus\Data\logs\RVN00017.log. Error -530.

and octopus server dies. currently our licensed octopus server is down.

for anyone else looking to:

Raven.Smuggler.exe in http://localhost:10930/ c:…file.octobak

Raven.Smuggler.exe out http://localhost:10930/ c:…file.octobak

was able to get back the site from backup…

Hi Leblanc,

If you are available, I can try to help work this out on Skype - my Skype name is paulstovell.


Thanks for the update Leblanc, glad you got it working. Octopus automatically creates backups under a Backup folder every few hours out of the box, but yes, the restore currently requires a command line call.

Do you have any thoughts on what might have caused the error? Was the server unexpectedly shut down for example, or did the Octopus service crash?


Octopus service would start and turn off right away. (main error above)
Esent was the source in eventlog

what saved the day was the ability to create a new instance of ravendb Data folder through the octopus admin tool.

once the service came back up I was able to import from backup.