Jira

rudelgurke

rudelgurke

Hello,

since there's a request on wantedports to create a port for Atlassian Jira, I'm working on it. First I've to see what Atlassian says and, overall, I guess FreeBSD won't make it onto their "supported OS" list.
I've quite some experience running it under Linux so getting it working on FreeBSD won't be a huge problem at all.
Open questions I've in mind - Jira has an application directory that can be put somewhere. Inside this directory there's a subdirectory "logs" which needs to be write-able - could be solved using a symlink.
Additionally it needs a "data dir" where everything else gets stored. Any idea what the best possible location might be ? Something like /var/db/jira has been cooking in my mind.

Side note - once Jira - if Atlassian has no problems with it - is ported getting other Atlassian products like Confluence, Bamboo, Bitbucket - where I've experience too - ported should be no problem at all.
 
We've been running Jira 7.1 and Confluence 5.10 on FreeBSD 10.3 for the last year (well only updated to 10.3 from 10.2 recently) using mysql as the backend and seems to work well. Had to make one change to the Jira code so that it would accept OpenJRE as a valid JRE (as I think it check for Oracle in the version string). Also had a bit to trouble getting a mysql connector working for confluence but I think that was down to me rushing through and not following their instructions correctly.

But both run without issue that I can see. I made rc scripts for them which I'm happy to post (first rc scripts I've made so may not be correct but they work for us). We run them under a jira user. Our directory structure is roughly

Code: 
/jira
|
|-- confluence
|   |
|   |-- confluence5
|   |-- data
|
|-- jira
    |
    |-- jira7
    |-- data

This is on zfs. It's the first FreeBSD system I've set up so may not be the best directory structure but wanted to keep everything jira-related in one place rather than splitting it across db and log direectories in var. I've done a similar setup for mysql (all in /mysql) also although this caused an issue when upgrading to 10.3 as it reverted the mysql rc script to point to /var/db (I think) so had to change that back.

But if Atlassian allow, a specific FreeBSD package/port would be really useful.

Thanks

Russell
 
I'm sorry if I've bumped a year old thread, but I came here looking for a simple rc script for Jira.

Since it's not been posted in this thread, I threw one together quickly and am sharing it here for future users. This is the first time I've created an rc script, so I apologize in advance for any errors - it works on my end.

Code: 
#!/bin/sh
#
# PROVIDE: jira
# REQUIRE: mysql nginx
# KEYWORD:

. /etc/rc.subr

name="jira"
rcvar="${name}_enable"

start_cmd="${name}_start"
stop_cmd="${name}_stop"
restart_cmd="${name}_restart"

load_rc_config $name
: ${jira_enable:=no}
: ${jira_user:="jira"}

jira_start() {
        su -l "${jira_user}" -c "exec ./atlassian-jira-software-7.13.0-standalone/bin/start-jira.sh"
}

jira_stop() {
        su -l "${jira_user}" -c "exec ./atlassian-jira-software-7.13.0-standalone/bin/stop-jira.sh"
}

jira_restart() {
        jira_stop
        jira_start
}

run_rc_command "$1"

It makes assumptions about the directory structure and the fact that the jira user is already created, but those are fairly straightforward to customize.

Credits to this thread for the base
 
Just curious what happened to this port attempt. Either way Jira is not FOSS. Jira server starter licenses (10 people or less) are currently $10 U.S. per year. Not too bad actually. They even donate the proceeds to charity.
 
Here's the reply I got from one would-be-maintainer over a year ago:

"Also the port(s) have been restricted to prevent source and binary
re-distribution because Atlassian - apart their marketing - didn't give
a clear reply here."

Here's the high-level build instructions from the same person:

"Basically it was done like:

- download the required tgz archive
- extract into ${PREFIX}/libexec
- symlink ${PREFIX}/libexec/atlassian-jira-1.2.3/conf to ${PREFIX}/etc/jira
- create a symlink to ${PREFIX}/libexec/jira
- add a init script that calls ${PREFIX}/libexec/jira/bin/server.sh

Additionally I've pre-configured the datadir to /var/db/jira"

I've not heard from this person since then. For the time being, it's safe to assume, if you want the Atlassian stack (not just Jira) - you're on your own to port it and maintain it.
 
You must log in or register to reply here.
Back
Top