Blog ·
Jira maintenance schedule: routine work that shows up on time
A Jira maintenance schedule is a list of recurring jobs, each with a cadence, an owner and a ticket that appears in Jira on the day the job is due. Jira has no maintenance calendar of its own, so you build one from three parts: an inventory of what needs doing and how often, a ticket shape that every job shares, and a way to create those tickets on time without anyone remembering to. This guide walks through all three, then covers the reports that tell you whether the schedule is actually being kept.
Statements about Jira's built-in features are as of October 2026, from Atlassian's own documentation.
What counts as maintenance work #
Maintenance is the work that keeps something running and never finishes for good. It comes back on a cycle. Some typical examples, by team:
| Team | Daily or weekly | Monthly | Quarterly or yearly |
|---|---|---|---|
| IT and infrastructure | Backup checks, disk and queue alerts review | Patch cycle, certificate expiry check | Access review, disaster recovery test |
| Software | Dependency update review, flaky test triage | Log retention and cleanup | Library upgrades, dead feature removal |
| Facilities and equipment | Safety walk, cleaning checklist | Equipment inspection, fire extinguisher check | Servicing contracts, HVAC service |
| Content and web | Broken link check | Analytics review, plugin updates | Content audit, legal page review |
What these jobs share is that nobody forgets them in the first month. They get forgotten in the seventh month, when the person who set them up has moved on and the reminder lived in their calendar. A maintenance schedule in Jira exists to make that failure visible, or better, impossible.
Step 1: build the maintenance inventory #
Before touching Jira, write the list. A spreadsheet is fine for this step. For each job, fill in five columns:
- The job, named the way you will search for it later:
Quarterly access review, notReview. - The cadence: every weekday, every Monday, the first of the month, the last Friday of the month, every three months.
- The owner: one person or one role who answers for it.
- How long it has to get done: a backup check is due the same day, a quarterly review can take two weeks.
- What done means: a checklist, a signed off report, a screenshot, a link.
The fourth column becomes the due date. The fifth becomes the description or the subtasks. Most teams find ten to thirty jobs once they sit down and list them, and two or three that nobody was doing at all.
Step 2: decide on one project and one ticket shape #
Put maintenance tickets in the project of the team that does the work, not in a separate maintenance project nobody looks at. The board the team already opens every morning is the board where the job will be seen.
Give every maintenance ticket the same shape:
- Summary with the job and the period, such as
Fire extinguisher check October 2026, so a search for one month finds it. - Label
maintenance, plus one for the job, so every run of every job is one filter away. - Assignee from the owner column.
- Due date from the window column.
- Description holding the steps, or subtasks when different people do different parts.
A checklist in the description is enough for a one person job. When the steps belong to different people or need to be tracked on the board, use subtasks. Our Jira checklist template guide compares the places steps can live, and Jira subtask template covers getting the same subtasks every time.
Step 3: create the tickets on schedule #
This is the part that decides whether the schedule survives. There are three ways to do it, and they suit different sizes of list.
| Clone by hand | Jira Automation scheduled rules | A scheduling app | |
|---|---|---|---|
| Who creates the ticket | A person, each cycle | A rule | A rule in the project |
| "Last Friday of the month" | Whoever remembers | Needs a cron expression | Chosen in the form |
| Weekends and holidays | Whoever notices | Up to the rule's logic | Set once for the project |
| Thirty jobs | Thirty reminders | Thirty rules | Thirty rules on one page |
| Cost | Time | Counts toward your plan's automation usage | App licence |
Clone by hand #
Keep one finished ticket per job as a model and clone it when the cycle comes round. It is free and needs no setup. It is also the method that fails silently: a missed clone leaves no trace, because there is no ticket to be overdue.
Jira Automation scheduled rules #
Jira Automation can create a ticket on a timer. Atlassian's trigger reference says the Scheduled trigger can "run the flow at a fixed rate (for example, every 7 days), or use a Cron expression for more complex schedules." A daily or weekly job fits the fixed rate. The first Monday of the month, the last Friday, or a quarterly date needs cron, and our Jira cron expression guide covers the syntax and the schedules it cannot express. Each run counts toward your plan's automation usage, which Jira Automation pricing explains.
Two things to plan for with a long list. First, every job is a separate rule, so a schedule of thirty jobs is thirty rules to keep track of alongside every other rule on the site. Second, the same Atlassian page warns:
Scheduled flows that reach a Failure status for 10 consecutive executions will disable automatically.
On a quarterly job, ten failures is more than two years. A maintenance rule that switched itself off looks exactly like a quiet quarter, so check each rule's audit log as part of the review in step 4.
A scheduling app #
Recurring Work Items for Jira is our app, and a list of maintenance jobs is the work it was built for:
- Turn the app on for the team's project and choose who may schedule.
- Set the working week and add the public holidays, so a job that falls on a day off is handled the way you choose instead of landing when nobody is in.
- Add a rule per job. The schedule options include every day, chosen weekdays, a date in the month, a weekday position such as the last Friday, and an interval such as every 3 months.
- Name each ticket with tokens so the period is in the title, for example
Fire extinguisher check {{now.monthName}} {{now.year}}. - Set the assignee, labels, subtasks and a due date as an offset from the run.
It uses no Jira Automation quota, because the scheduling is the app's own. Every run, created or failed, is written to the rule's history with the reason, and the rules table shows each rule's next run and health on one page. Rules only create tickets in the project they live in. It is free for up to 10 users.
Step 4: report on whether the schedule is kept #
A schedule is only as good as the evidence that it ran. Three saved filters cover most questions. Replace OPS with your project key.
- What is overdue right now?
project = OPS AND labels = maintenance AND due < now() AND statusCategory != Done ORDER BY due ASC - Did every run of a job happen this year?
project = OPS AND labels = access-review AND created >= startOfYear() ORDER BY created ASC - What is due this week?
project = OPS AND labels = maintenance AND due >= startOfWeek() AND due <= endOfWeek()
Put the first filter on the team's dashboard and subscribe the owner to it weekly. The second is the one an auditor or a manager asks for: four quarterly tickets this year, four resolved, each with its date. A gap in that list is a missed cycle, which is the one thing a calendar reminder can never show you. Our Jira dashboard examples and Jira weekly report guides show how to turn filters like these into something people read.
Once a quarter, review the inventory itself:
- Is every job still needed? Retire the ones that are not, rather than resolving tickets nobody acts on.
- Is the cadence right? A check that has found nothing for a year may move from weekly to monthly. One that keeps finding problems may need to move the other way.
- Does every job still have an owner who is still on the team?
Common mistakes #
- One giant recurring ticket for all maintenance. It is never done, so it is never overdue, so it tells you nothing. One ticket per job per cycle.
- Tickets created far in advance. Creating a year of monthly tickets in January fills the backlog with noise and every change to the job means editing twelve tickets. Create each one shortly before it is due.
- No due date. A ticket without one never appears in the overdue filter, and the overdue filter is the whole alarm.
- Maintenance in a project nobody opens. The schedule runs, the tickets appear, nobody sees them. Use the team's own board.
When the manual way is enough #
If your list is five jobs, one person owns all of them, and they are reliably there every cycle, cloning a model ticket is enough. Keep the label and the due dates anyway, so the overdue filter works. A scheduled approach earns its place when the list grows past what one person can remember, when the jobs are shared across a team, when the cadence follows the calendar in awkward ways (last working day, first Monday, every quarter), or when you need a year of evidence that the work happened.
For specific maintenance routines in more depth, see Jira patch management, Back up Jira Cloud and Jira hygiene. For the broader comparison of ways to repeat work, see Jira recurring tasks.
Frequently asked questions #
Does Jira have a maintenance schedule feature? No. Jira tracks the work. The schedule comes from how you create the tickets: by hand, with Jira Automation scheduled rules, or with a scheduling app.
Should I create all of the year's maintenance tickets at once? Usually not. Tickets created long before they are due clutter the backlog and go stale when the job changes. Create each one close to its due date.
How do I schedule a quarterly maintenance task in Jira? With Jira Automation, use a Scheduled rule with a cron expression for the quarter's date. With Recurring Work Items for Jira, choose a monthly schedule with an interval of 3 and set the due date as an offset from the run.