How do I syncronize with the JIRA DVCS Connector without the Bitbucket service?

[Kasper Bakmann]

JIRA is installed on a server with no public domain name, so the problem would be to connect from Bitbucket to JIRA. Could I set up JIRA to syncronize on a specified time interval ?

Same issue for GitHub users as discussed here:

https://answers.atlassian.com/questions/57829/setup-dvcs-force-sync-as-a-service

[Kasper Bakmann · Accepted solution]

A newer version of JIRA (5.1.3) helps with this issue and it is no longer a problem.!

[Kasper Bakmann]

A newer version of JIRA (5.1.3) helps with this issue and it is no longer a problem.!

[Kasper Bakmann]

Thanks Norman for your effort and time.. I was hoping for an easier solution, for now it is cheaper for us to go into JIRA and press Force sync rather than implementing this.

[Norman Abramovitz]

I was attempting to look up the steps but I think the connector documentation may be sufficient.

Ideally you want

loop

get log information since the last time you got log information

update jira with the log information

end

record the last update to be used as the start time.

So

I can see two approaches.

clone the repository on your private network

install the dvcs connector on the clone repository

Now update a file in a different repository, commit and push

now do a pull operation on the newly cloned repository

Now see if Jira was updated.

The other approach is

build a separate java application based upon the dvcs plugin code that updates Jira. This gives you the all the calls necessary to update Jira. Lets say this application reads a file.

Now use the bit bucket command to get log information and format the results to be input to your jira update program.

Now you can build a script that pulls from bitbucket, creates the log input file and then updates jira.

[Kasper Bakmann]

Ok .. i am willing to try .. do you have any links or guides how to do it ?

[Norman Abramovitz]

Since you are willing to code, the subversion plugin does a pull operation from the jira side based upon a time interval. This gives you a framework for the service. You then integrate/code the DVCS plugin code that updates Jira into subversion service and then change the subversion code that pulls the history information to use the bitbucket commands for log history.

[Kasper Bakmann]

I am not using Subversion but HG Mercurial, so i do not think it is an option.

[Norman Abramovitz]

My internet connection went out. You are replacing the subversion code with the bitbucket/mercurial code.

[Norman Abramovitz]

It is not another link in the chain, it is more like another developer and that developer is Jira.

[Norman Abramovitz]

Are you willing to do some medium level of coding? If not, then the shadow repository mentioned in the referenced question should work for you.

[Kasper Bakmann]

Im willing to do som medium coding .. but where do i start?

The shadow repository seems a bit overkill .. the reason for using Bitbucket is that it is outside our firewall .. and bringing in another link in the chain with shadow repository is not an option.