Forums

Articles
Create
cancel
Showing results for 
Search instead for 
Did you mean: 

Running JIRA and Stash with the same MySQL Database Server

Veracode
Contributor
November 21, 2013

We are running JIRA 6.0.3 and Stash 2.6. Each connects to its own MySQL database which is on a separate machine.

We have a database server where we run the MySQL databases for our Atlassian products.

All of this is duplicated for DEV vs PROD environments.

We would like to upgrade Stash, but the documentation for Stash 2.7 indicates that we must change the settings of the MySQL database server in a way that contradicts the settings required for JIRA.

  • Stash
    • binlog_format=mixed
  • JIRA
    • binlog_format=row

I asked Atlassian about this, and they indicated that "JIRA and Stash behave this way", and that I should see if I could get help via asking the question at answers.atlassian.com

We would prefer to run our databases on the same database server, rather than deploy a separate machine for each individual database. I guess I am asking if there is a way to get around this problem without placing the JIRA and Stash databases on separate machines (i.e. separate MySQL installations).

If we determine that separate servers are required then should we take the hit now and place the Confluence database on a separate machine as well in case this type of conflict happens between Confluence and either JIRA or Stash in the future?

Any suggestions?

References:

https://confluence.atlassian.com/display/JIRA/JIRA+Cannot+Create+Issues+when+Using+MySQL+with+Binary+Logging

https://confluence.atlassian.com/display/STASHKB/Stash+2.7+fails+to+start+with+MySQL+when+binary+logging+is+enabled

1 answer

1 accepted

1 vote
Answer accepted
ThiagoBomfim
Atlassian Team
Atlassian Team members are employees working across the company in a wide variety of roles.
November 21, 2013

Hi Veracode,

In the official Connecting Stash to MySQL document we state:

If MySQL is using binary logging, configure the database to use a binary logging format of either MIXED or ROW.

I have tweaked that KB article to match our official document - so we will avoid conflict of information. Hence, you should be fine with the current "binlog_format=row" on the JIRA database.

Do open a ticket with us in case you have problem and thanks for pointing this out.

Best regards,

Thiago Bomfim

Atlassian Support - Stash

Suggest an answer

Log in or Sign up to answer