This page last changed on Oct 14, 2010 by alui.

It is possible to have files in your repository whose names differ only in case, e.g. Foo.java and foo.java. Hence, your database will need to use rules for comparing string values which recognise that upper and lower case letters are different, that is, a 'case sensitive collation'.

If your database was not created according to these instructions: Migrating to an External Database, then you may be using a case-insensitive and/or non-UTF8 collation. If so, FishEye will display the following message at the bottom of your screen:

"Your database is not using a case sensitive UTF8 encoding for character fields."

The following sections provide instructions for changing your database collation for each database type supported for FishEye/Crucible.

On this page:

MySQL

 

Please take a backup of your database before changing its collation.

To change your collation to utf8_bin you need to change your database's default collation, but as this only affects newly created tables you will also need to change the collation on the table for which case sensitivity is critical.

Change your Database's Collation

Use the ALTER DATABASE command, as follows:

alter database character set utf8 collate utf8_bin;

Change Collation for CRU_STORED_PATH Table

Use the ALTER TABLE command, as follows:

alter table cru_stored_path convert to character set utf8 collate utf8_bin;

PostgreSQL

 

Please take a backup of your database before changing its collation.

If you have created your PostgreSQL database with the incorrect encoding, you will need to dump your database, drop it, create a new database with the correct encoding and reload your data.

You can do this via the standard database migration procedure – instead of migrating from HSQLDB to PostgreSQL, you are migrating from a PostgreSQL DB with the incorrect encoding to one created with the correct encoding.

Document generated by Confluence on Oct 21, 2010 00:42