Database collation. Case sensitive or insensitive

Question

Getting ready to upgrade from 4.3.7 to 5.7. Checking our DB collation, we seem to be running CASE INSENSITIVE when the 5.7 doco states your DB must be CASE SENSITIVE.  Checking back through all the documentation, it says the DB collation should be CASE SENSITIVE.

We have  been running this way for over 9 years without an issue on various versions of Confluence. Anybody know the reason they require CASE SENSITIVE?

Ingo-Stefan Schilling · Answer

Honestly, one of the annoying things was case sensitivity in Confluence is the fact that labels can be found only if written in right cases, so is content.

Since UTF-8 and case sensitivity have nothing to-do with each other, this can't be the answer nor reason.

@Atlassian it would be great to get a profound answer - and to be honest, it would be great if this case sensitivity thing can be removed - we would be at least grate- and thankful since it would avoid hours of explaining users why it is as it is... which consequently would optimize TCO of Confluence dramatically.

I opened a feature (improvement) request https://jira.atlassian.com/browse/CONF-39019 so if someone else feels as if there should be changed something, feel free to vote there

Mike Welch · Answer

Is this collation verification new to 5.7?  We have been using case insensitive since 3.x. We are an integrated Jira/Confluence shop and notice that Jira requires case insensitive, unlike Confluence. Why the difference?

Septa Cahyadiputra · Answer

Confluence 5.7 require you to use UTF 8 and case sensitive collation. During the upgrade process to 5.7 Confluence will check the collation of your database, and if it does not met the requirement, Confluence will failed the upgrade process.

Hope the above information clarify your doubts on this issue.

Cheers,
Septa Cahyadiputra