Faster way to replicate a clone from the archive

[Mark de Bont]

Is there a faster way to get a bare GIT repository onto your Bamboo agents?

Our archive 5+GB and takes more then half an hour to clone over to the agent. I've enabled repository caching and am using shallow clones on the agent but occasionally it starts (re)cloning the complete archive .

This takes more then half an hour and then the jobs times out.

I was thinking of creating a temporary copy on a file share to have some sort of baseline to start from.

[Marcin Oles]

Hi Mark,

which VCS are you using? If it is Git, is option "Use shallow clones" enabled for the repository? It causes Bamboo and agents to only checkout most-recent commits, skipping as much history as possible. The option is available for multiple Git-related repository types in Bamboo, including Git, Bitbucket Server and Bitbucket Cloud. See https://confluence.atlassian.com/display/BAMBOO/Git for more details.

Regards,
Marcin

[Mark de Bont]

Hi Marcin,

Yes it's a GIT archive and I am already using shallow clones.

Even though I've marked this, it appears it is still downloading the complete archive ;-s

Cheers Mark,

P.s. I've edited my question to clarify it's a GIT archive and using shallow clones.

[Mark de Bont]

When I do a shallow clone with depth 1 on my workstation the .git repository only is 237MB.... and not 5BB