Archiving of jira projects

[Rahul Aich [Nagra]]

Hi All

We are observing a rapid growth in our jira instance with almost 40% increase in number of issues every year.

What are the archiving options we have?

Note:

- we cannot delete old issues as we want to retain them for search

- we cannot hide the project by restrciting the permissions as we want them for search

- making the project read-only serves little as it does not improve performance of a big instance

- project import on to another server is also not an option as it requires maintaining configurations in sync with source and target; moreover, on huge jira instances xml backups can go over 150MB and sometimes cannot be taken for memory issues.

My assumption is that with growth of your instance the performance of instance degrades and hence the need to archive.

Rahul

[Nic Brough -Adaptavist- · Accepted solution]

You've neatly ruled out all of the current options in your question.

I've had discussions with Atlassian people about archiving in the past, and I know they are actively working on something that will help you with it. But I'm afraid I don't have any timescales or details that I know are accurate enough to be worth sharing.

[Peter Illig]

Hi all,

last answers in this topic was almost 4 years ago. So I'd like to repeat the question: is there a possibility to archive projects?

Archive from our point of view, means that a specific project has a frozen status, it is not possible to work on it anymore, no login, nothing. But in case of audit, we can reopen it to a readonly-mode.

Archive is important to relieve the database, for clarity but also to be able to handle audits.

Thanks,

Peter

[Nic Brough -Adaptavist-]

No change.

But your description of an archive is one of the standard approaches already.

Create a permission scheme called "Archived".  Grant only "Browse project" permission to one role.  Apply it to the projects you want archived.  Remove all users (except maybe admins) from that one role.

[Zitrax]

Jira Software 7.10 added project archiving, however only for the Data Center version, see: https://confluence.atlassian.com/jirasoftware/jira-software-7-10-x-release-notes-949226697.html#JiraSoftware7.10.xreleasenotes-archiving

[Peter Illig]

Is there any update from Atlassian side? The archive function was rolled out to the JIRA cloud customers but especially for large companies that host JIRA on their own server farm (due to regulations), we definetly need a solution for archiving.

Please let us know, when this important function is going to be released. Thank you.

[Nic Brough -Adaptavist-]

Archiving is in Jira 7.10 and above, but data centre only, I don't think Atlassian plan to bring it to server.  I suspect the thinking is that if you're big enough to be thinking of archiving, you probably should be on Data Centre anyway

[Peter Illig]

Maybe. But unfortunately, reality is slightly different: finance- and insurance companies as well as government are not allowed to use Data Centre solutions due to data security reasons. So specially for the large customers like us, the data Centre will never be an option.

[Nic Brough -Adaptavist-]

Ah, I think I know what's wrong here.

You are mixing up Cloud and Data Center.  They are totally different things.  Whilst Cloud has problems for some organisations (location of data, access by people outside the organisation, scale, etc), Data Centre is heavily used by Finance, Insurance, Government, and even Security services.  DC is a cluster of servers intended for scale and high availability and can be installed anywhere you want it.

The reality is the complete opposite - companies with data security reasons often need Data Centre for scale, and a significant proportion of DC customers are financial!

[Remie Bolte]

@Rahul: can you describe the definition of "Archive" if you still want to be able to search through issues / projects and keep watches on those issues? That sounds like active projects to me.

What is your goal? What kind of performance issues are you talking about? Is is server level (CPU, memory, storage) or application level (search, filters, agile board, gadgets)?

[Nic Brough -Adaptavist-]

I suspect "archive" is similar to what I've been asked to look into loads of times - a whole project, no updates going on at all, should be read-only, but you still need to search for it.

[Rahul Aich [Nagra]]

@Remie For us definition of archive is - an old project which is no longer active (no new bug submitted and no activity on it) however, that project issues is often queried to find out fix implemented and their fix versions. So, we cannot delete them nor can we take xml backup of it as it needs to be accessed by users.

Moreover, if we leave them as it is on JIRA then the drawbacks are

1. Half a million issues will cause slowness when searching. Higher indexing times.

2. Hundreds of old projects means current active project owners will find it difficult to find their active project from the project list

3. Create issue button will take time to load since it will have to load all the projects and its issue type scheme.

4. generally, after 200,000 issues Atlassian recommends to split instances as it causes performance issues (generally while accessing and updating issues)

5. Currently Agile boards are taking ages to load and we suspect it is because of high number of issues we have.

Rahul

[Nic Brough -Adaptavist-]

One minor point - the 200,000 limit was a recommendation for Jira 3. Although Atlassian haven't really announced another number, Jira 4 and 5 can happily handle a lot more, and you can get into millions if you're willing to put some effort in on decent hardware and proper application tuning.

[Anand Unadkat]

Hi Rahul,

There is no way that you will be to archive projects. To make is more obvious though, you can categorise projects. So the projects that you want to archive, create a category for it, for example 'Archived' and add those projects to that category. Will not make any changes to your performance.

Regards

Anand