How to make Jira importer to accept a file of 5GB?

[Pradeep Pamidipalli]

Hi,

I am in a process of evaluating Jira for my organization. We are currently using Trac for tracking bugs. Following is the information realted to the Trac environment:

Trac 0.11.7

DB - PostgreSQL 9.1.2

Jira - 4.4.4

I have tried with a similar setup on my local. Read about the Importer plugin of Jira and upgraded Trac to 0.12.2 version to have it supported. Am able to import sample projects created in Trac.

Now when am trying to import the actual trac data which is of 5 GB, the importer plugin does not allow me to ponit a file which is more than 2GB. I have the following questions in this regard:

Q1. Is there a way to make the importer plugin accept of file of size more than 2GB?

Q2. Will the importer plugin accept a file other than a .Zip file for import?

My trac folder has the attachments folder which is the one causing the problem. The attachments folder is 4.5GB in size.

Q3. Is there a way that I can import my attachments by pointing it apart from the importer?

Would be great if I can get the answers in quick time.

Thanks,

Pradeep

[Wojciech Seliga · Accepted solution]

JIRA Importer Plugin uses underneath the standard JIRA temporary attachment functionality which is currently limited to files 2GB in size. So there is no easy workaround currently for your problem.

See https://studio.atlassian.com/browse/JIM-601 and https://jira.atlassian.com/browse/JRA-25858 for related issues.

Ad Q2. Trac importers expects a zip file.

Ad Q3. Unfortunately no.

The solution to your problem would be to support bigger fies than 2GB, which I hope will be implemented soon (it's XXI century after all).

If you don't mind changing a source code of JIM and rebuilding it by yourself, then you could read the attachments from another place, but modifying copyAttachment method in com.atlassian.jira.plugins.importer.imports.trac.transformer.AttachmentTransformer class.

Then your zip file will supposedly go far below 2GB.

You can get JIM source code with JIRA source distribution. It can easily built as any other regular JIRA plugin using Atlassian Plugin SDK.

[Pradeep Pamidipalli]

Thanks for your answer Wojciech Seliga. I have found a work around for my problem.

With Trac 0.12.2 version, we have a plugin that can be used to delete tickets. I have made my huge zip file into two now with the following steps:

Step #1: Create two empty DBs - t1 & t2 in postgresql

Step #2: Import trac.sql data into both the DBs

Step #3: Create two new trac environments

Step #4: Change trac.ini files to point to t1 & t2 correspondingly

Step #5: Run the deletetickets.bat file to delete new tickets in t1 and old tickets in t2 environments

Step #6: zip the contents of t1 and t2 seperatly

Step #7: Login to Jira as an administrator

Step #8: Create a new project and Import t1 by poiniting the t1.zip file

Step #9: Import t2 by poiniting t2.zip into the same Jira project

Step #10: Verify if all the tickets and corresponding attachments are imported successfully.

Untill we have the improvement ticket (https://jira.atlassian.com/browse/JRA-25858) closed, we will use the work around.

Thanks,

Pradeep

[Jack Olsen]

To repair corrupted zip files you need to use programmed instrument. I would recommend the next one Zip file repair

[Pradeep Pamidipalli]

A quick update on this !!!

I have used maximum compression and brought down the file size to 2.8GB. Is there a way to tweek the Jira settings to make it accept my 2.8GB file in the Importer plugin page?