cloud edition Exclusive locks using replicated repo

[Andrew Carvalho]

I am using Plastic Cloud  and have recently moved from connecting directly to the cloud to a local replication. There are times when I am working when I don't have internet (I commute a good portion some days and usually catch up on work) or when my internet connection is spotty.

It's something I have wanted to do for a while but help off on due to reliance on exclusive locks. However, I decided to investigate if there was a solution to this and came across this blog post and how the lock.conf file can be placed beside the server executable to redirect lock requests and checks to another server.

I've created the lock.conf file but am unsure how to point it to our cloud.  I have tried

rep:[out_repo] lockserver:[our_org]@cloud:8084

but it does not work as I'm not sure if the server will automatically expand the @cloud shortcut.  I've also tried port 8086 as it was mentioned in the Cloud guide. Though the guide mentions lock.conf being hot-reloaded, I restarted the service just in case to make sure that wasn't the issue as well.

What would be the appropriate way to setup my local lock.conf to use plastic cloud as the lockserver? Is this not currently possible? If not I will likely move back to a centralized system (which had no issues other than needing that pesky internet connection). I'm aware the locks won't work if I am working without an internet connection but as I am usually connected and really just want the distribution for the odd time I'm not, I'm okay with this.

Bonus question: if I am redirecting my locks to a different server, do I still need to list all the patterns in my local server?  I have them in the file currently but wasn't sure if it was necessary.

 

Andrew

[calbzam]

Hi Andrew,

I'm afraid that distributed locks for the cloud is not supported. It's in our roadmap to include it (we have some good ideas) but I'm afraid this feature is not still scheduled. All you feedback is very welcome.

If you want to use locks (I guess with Gluon?), we recommend you to follow a centralized workflow pointing your workspaces directly to the central server.

Regards,

Carlos.