Why so many issues in backlog?

[Jirong Hu]

I've migrated thousands of issues from Rally to JIRA, now seeing many in the JIRA backlog.

I see the backlog query is showing like this:

project = CVE AND status in ("10100", "3", "11600", "10002", "11603") AND (cf[10004] is EMPTY OR cf[10004] not in openSprints()) AND issuetype in standardIssueTypes() AND issuetype != "10000" AND NOT status in ("11600") ORDER BY Rank ASC

From https://rbcjira.devfg.rbc.com/rest/api/2/field, I see cf[10004] is Sprint. How can I find out what's the name for the ids in the status?

[Jamie Echlin _ScriptRunner - The Adaptavist Group_]

That query is generated by jira. The contents of the query is nothing to do with us. We're just surfacing it so that you can view the backlog issues in the issue navigator. That is the actual query jira is using to show the issues in the backlog on the planning board.

> The weird thing is when I look at our JIRA PROD server, the query for Backlog is simple as: 

project = CVE and sprint is empty ORDER BY Rank ASC

You must have an older version of SR4Jira, this was changed as part of https://productsupport.adaptavist.com/browse/SRJIRA-3954?jql=project%20%3D%20SRJIRA%20and%20text%20~%20%22backlog%20query%22.

Although now I look I see there is a follow-up and a new issue: https://productsupport.adaptavist.com/browse/SRJIRA-3366. I haven't properly investigated that yet.

Either way, check the version of the plugin you have and see if either are before 5.6.8.

[Jirong Hu]

Hi Jamie

Thanks for your time to looking into this!

I think I am hitting a similar issue, but I still don't know how to fix it. The version is 5.6.11.2-p5 as shown below.

[Jirong Hu]

Current SR version in our JIRA PROD is v5.4.12, where I don't see the issue (but I haven't import my data yet).

Not sure if the problem in our JIRA DEV server is created by my import is my real concern.

[Nic Brough -Adaptavist-]

That's quite a complex query, and if it is viewing "too many" issues on the board, I would say there is something not quite right with the OR parts.

Decompose it, and try to simplify.  Do you really mean

• All issues where the Project is CVE where status is in (list) and Sprint is empty
• And
• All the issues where Sprint is closed and the issues are story level and of type X and they're not in a certain status

My best guess is that the second clause is grabbing vast swathes of issues from other projects.

I would simplify:

• Generally, a board does not exclude most of that.  You want to select all status, otherwise you will miss stuff.  Although you can tell your board to simply not have certain status included in a column, so it won't show, that's quite counter-intuitive
• You don't want open sprints only on a board, or you won't be able to plan
• You shouldn't be worrying about issue type so much.  Issue is of a few types, maybe, but I wouldn't bother with excluding sub-tasks in the board filter, the board understands they're not stories and they can be useful in a sprint, so instead think of hiding them later, giving you a more generally useful board

I would use

• "Project = CVE" as the board filter
• Possibly a clause for issue type (leaving sub-tasks on the board)
• If I was certain there were status that should not happen in my sprint, then I would first drop them from the board by removing them from the columns they were in, and then think about amending my workflow so it simply no longer uses them
• Have a "quick filter" on the board for "issueType is not in subtaskTypes()" - this enables you to quickly toggle their display on and off

[Jirong Hu]

Hi Nic

Thank you so much for the detailed explain. I think I missed something here.

See the attached screenshot. I simply access the Backlog from the menu left. The query was showing when I click the icon looks like a CPU. I was assuming the backlog is generated by a fix query in JIRA, but seems like the query is generated by looking at issues in Backlog (the other way round)?

[Nic Brough -Adaptavist-]

The backlog is

• Everything selected by the board filter
• Minus done issues
• Minus sub-tasks
• Minus things in an active sprint

The backlog absolutely is generated by the board filter, not the other way around.  Your filter is complex and long.

Can you try going to "board configuration" to see what we're really looking at?  Should be some ellipses (...) near the top right

[Jirong Hu]

Is this what you are looking for? The filter query is project = CVE ORDER BY Rank ASC

I mean I didn't create that complex query. That query shows when I click the "CPU like" icon. 

[Jirong Hu]

See my 1st screenshot, the 2nd issue is belong to "Sprint 2", but it's still in backlog? So now the question is back to how this backlog query is generated?

[Jirong Hu]

If I change the query to project = CVE AND (cf[10004] is EMPTY and fixVersion is empty and status not in ("Accepted","User Accepted", "Done") ) ORDER BY Rank ASC, then I get correct result.

Just don't where does that query comes from.

[Nic Brough -Adaptavist-]

I think you seem to be looking at two separate boards.

[Jirong Hu]

I am looking at the same board. Now when I create a new board, everything is correct. I compare the config of these two boards, can't see any difference.

The simple question is: for this Backlog menu, how does its query created? It's system default or what?

[Nic Brough -Adaptavist-]

Ah, I think I know part of the answer.

That blue icon you are clicking is being injected by one of your apps, and it's doing heck-knows-what.  None of the Cloud or Server (or the DC) installs I've got access to have it.

I think we need to find out where that is coming from and what it is supposed to be doing.

[Jirong Hu]

I spin up a new JIRA trial, you are right, base JIRA doesn’t have this icon. I think it’s from plugin “Adaptavist ScriptRunner for JIRA”. Let me contact my JIRA admin and find out

[Nic Brough -Adaptavist-]

Ah, that has just kicked me in the memory!  Thank you so much.

It absolutely is coming from ScriptRunner.  It is there to try to give you the option to see the sprint, or backlog, that is currently in front of you.

But I think the complexity and an oversight is misleading us here.  I have certainly mis-read one of your screenshots.

I think the second thing I've missed is the Sprint field.   If an issue is in a sprint, it will be grouped within that sprint, and will not be in the backlog section as shown in your screenshot.  There are two ways to get a display like what you have there and you may be suffering one or both!

First, you have "backlog" named (with 1442, then the ScriptRunner icon) on the top line.  Whilst this looks like a backlog, you might get that display if you have a sprint named "Backlog".  I do not know if you have done that, but if you have, please don't.  "Holding sprints" break all sorts of things.

Second, if you look at the second line in the display of the three issues, they start with none/sprint 2/sprint 2.   This does not look like the Sprint, it looks like a (probably custom) field that holds some options (obviously, one of which is "Sprint 2").  But not the actual Sprint field.   This makes me think that when we go back to your original question, there's a problem.  You say "cf[10004] is Sprint" which is accurate, but also misleading (by Jira, not you!).  It would be better to say "cf[10004] is a field called Sprint which is just a text or select field, not the actual Sprint"

So, more digging needed.  Two questions

Can you go to Admin -> Issues -> Custom Fields and use ctrl-f to look for "Sprint".  How many hits do you get?  (I'm looking for an answer of "3", but it might be translated - the more complete search is that I am looking to check that you have one, single, field called "Sprint", which is "locked", has a type of "Jira Sprint Field" and a description of "Jira Software sprint field". 

Can you go to Issues -> Search, and run the advanced search of "Sprint = Backlog"?  Do you get any results, or an error?

[Jirong Hu]

Hi Nic

Thank you so much for your input. Unfortunately the answer to all your questions are no.

1. There is no Sprint called Backlog.

2. There is only 1 Sprint in the custom field, and the id is 10004.

3. There is no issue with  "Sprint = Backlog"

Please note: now when I create a new board "Jirong", the Backlog shows right. The query shows: 

project = CVE AND status in ("10100", "3", "11600", "10002", "11603") AND (cf[10004] is EMPTY OR cf[10004] not in openSprints()) AND issuetype in standardIssueTypes() AND issuetype != "10000" AND NOT status in ("11600", "10002", "11603") ORDER BY Rank ASC

Compare to the query in the bad board, the difference is later added two more status, and If I use this new query to replace the one in the bad board, the result is the same. So the question is still: how this query is generated?

[Jirong Hu]

Another issue I found is the Sprint Report doesn't shows a chart, not sure if it's related. It's even the same in the newly created board "Jirong". Also I couldn't change the Sprint Complete Date. This is how I migrate these issues from Rally.

1. Create all Sprints (maps to Iterations in Rally) with startDate, endDate.

2. Import all issues from Rally, map the Iteration Names to Sprint IDs.

3. Use a loop to Start and Complete all Sprints, using REST API. Because of this, the Sprint Complete Date shows the date I run the REST API. I tried to update it through REST API but doesn't work. That's why in the chart, the end date is this month.

4. Note that "Issue added to sprint after start time".

[Jirong Hu]

Confirm the icon is added by ScriptRunner. Search backlog in this page: https://scriptrunner.adaptavist.com/latest/jira/jql-functions.html

[Jirong Hu]

All right, I think the Sprint Report is right, since I added all issues in the same day so the chart can't show anything useful.

[Jirong Hu]

Finally I found the difference of the old board (wrong) and new board (right): the columns are difference, therefore the generated queries are different (in the status).

[Jirong Hu]

Just so let you know, all of the above I am working in our JIRA DEV server.

The weird thing is when I look at our JIRA PROD server, the query for Backlog is simple as: 

project = CVE and sprint is empty ORDER BY Rank ASC

[Jirong Hu]

I contacted my JIRA admin, he doesn't know. 

I notice if I move the columns around in the board, the query for the Backlog changes. Just don't know how it works.

[Jirong Hu]

I figured creating a new board won't fix this issue. The difference of the old and new board are the columns as shown here.