First production upgrade of Raven from 2198 to 2360

[Jason Dentler]

Oren & friends,

We're running build 2198 (from Jan 3rd) for our production website. We have 16,403 documents and 46 indexes. Every page on the website loads a document or queries several of them.

On a fairly regular basis, the site goes down as every Raven request times out.

Two questions:

What's the proper upgrade procedure?

Is this something that might be fixed by upgrading to 2360?

- J

[Oren Eini (Ayende Rahien)]

Well, 2198 is not a stable release, so a LOT of stuff has been updated.

I can't tell you more without additional data.

Upgrade procedure:

* Shutdown the server

* Backup the data files

* Delete the old binaries

* Copy the new binaries

* Start the server

--
You received this message because you are subscribed to the Google Groups "ravendb" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ravendb+u...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[Kijana Woodard]

Can this procedure be used for migrating from 2330  to 2.5? Is it generally applicable between raven versions?

[Jason Dentler]

What sorts of info should we collect? Memory consumption is high. 16+ GB for a 1.5 GB database. CPU sits at 0 for the raven.server process.

[Chris Marisic]

Yes. RavenDB data files are always safe for forward migration. Note that it's not necessarily possible to go down in versions if you haven't taken a backup prior to upgrading.

[Oren Eini (Ayende Rahien)]

Yes, but it isn't applicable bckward.

[Jason Dentler]

If our time-out problem continues, what sorts of information do you need to help us figure out the problem?

[Oren Eini (Ayende Rahien)]

The dump file, the debug logs.

[Jason Dentler]

So far, it hasn't happened since we upgraded to 2360. However, we do have the issue with that one multi-map/reduce index. In the other thread, I linked to a repo with my failing test.