Can we connect the JIRA6.4 to an empty database first while configuring and later we can offload all the data which we have in our actual database?

[Mrityunjay Kumar]

We are facing issues while connecting the database to JIRA6.4 after setup. We are planning the following.

1. Setup JIRA6.4 as per the process.
2. Connect the JIRA6.4 to an empty database which have the same schema.
3. After successful configuration we will offload the data present in the actual database to new database.
   
   I want to ask is this process feasible and effective or not.
   If someone have some other suggestions please share. Thanks a lot in advance. 

[MikeyS]

I don't know if the schema has changed between 6.2 and 6.4, but if you use the XML backup method, this should not matter. You are correct that connecting JIRA to the empty database will generate the correct schema for that version. You should then be able to tell JIRA to restore from your XML backup and it should map the data into the correct locations, even if the schema has changed. Attachment data will need to be handed separately.

[MikeyS]

That should be fine.  To prompt JIRA to ask for a new database you can just delete, or better yet rename the dbconfig.xml file from your JIRA home directory.  You can either transfer your data directly between the databases or you can dump it into an XML backup and have JIRA import the data into the new database using the data from the XML backup.

[Mrityunjay Kumar]

We had taken the XMLbackup for the earlier database of our instance. The backup size was more than 250MB and hence while importing some exceptions were being thrown. Hence we could not use it. Our actual database size was also around 2GB. Can you suggest why exceptions are being thrown? We looked for answers in atlassian and found that such large volume of data are not recommended to be imported through the XMLbackup. Is this correct or we are missing something here.

[MikeyS]

Okay, I think I may have misunderstood your original question. Are you trying to upgrade JIRA or just move the data to a separate but identical database? With regard to data size, that does sound like a decently large JIRA instance. You may have better luck getting a reliable XML backup if you take the backup during off hours or even downtime. If you are upgrading, you may not necessarily be able to reuse the same data from the database as the schema does occasionally change between versions. In that case, it would be best to use the XML backup to avoid having to directly modify the schema.

[Mrityunjay Kumar]

Hi Mikey, First thanks for responses. We are actually upgrading our JIRA from 6.2 to 6.4. As you are saying the schemas of database between the versions may change, does this apply here. I mean does the schema of JIRA6.2 different from JIRA6.4. If this is the case and if we are connecting our JIRA to an empty database in the start, it should create a new schema itself. We do not have to create a new schema altogether before connecting the empty database to JIRA6.4. If this can be done than how can we export our older data from the earlier database. This was my understanding as per the conversation and my need. Can you suggest the ways for this?