Migration tools for DB2 migration to MySQL

Question

Hi,

I am looking for migration tools for database migration from DB2 9.7 to MySQL 5.5. As DB2 is no longer supported for Confluence 5.x we decided to switch to MySQL. XML Backup is not an option as we have large datasets.

Has anybody experiences on this topic?

Thanks!

Nina

Deividi Luvison · Answer

Hey Nina, Based on that that bellow procedure: 1 - Generate a xmlbackup without attachments. 2 - Install a fresh Confluence on the same version as production. 3 - Make sure to install this new confluence with the same directory structure as production (if the is in c:\atlassian\ the target server needs to be exact the same). 4 - Once you see the input license page, shutdown Confluence. 5 - Follow the below guide so you can increase confluence memory as much as you can in the new installation server: https://confluence.atlassian.com/display/CONFKB/How+to+Fix+Out+of+Memory+Errors+by+Increasing+Available+Memory#HowtoFixOutofMemoryErrorsbyIncreasingAvailableMemory-Step2:IncreaseAvailableMemory 6 - After you do that start Confluence again and try to import the xml backup without attachments. 7 - Once you are done shutdown confluence again. 8 - Copy the attachments folder from production and drop it into your new directory. # Start Confluence and check if your attachments are there. # Profit! Let me know if that works. Thanks and Regards, David|Confluence Support Engineer

Alexey_Rjeutski__Polontech_ · Answer

Sometimes we have to do some kind of such migrations for our customers and found the workaround: we make a copy of instance with database etc into amazone ec2 server - create that one much larger for memory and with set of CPU cores - then make the backup there - then restore on the new configuration there - and then clone to normal instance. Even for huge instances that will cost ~10-20 dollars if do that fast.

Nina Daalmann_o3l4oz · Answer

Hi David,

thanks for the answer! We already tried the XML Backup but had to kill the process before even starting the export as CPU went up to 100% while rendering the pagetree. It is a large space with 88.233 pages.

Regards,

Nina