Jira Instance Cleanup Guide - Fields, Spaces, permissions, and tooling

aradu

👋 Hey there! I’m Alex, a Principal Engineer within Jira. I'm here to follow up on our very successful post around Jira Instance hygiene 🛀. This post will follow up with all of the changes in Jira cloud to help with this topic, as well as to address common questions in the comments - especially around what to clean, how to prioritise, and how to keep large sites fast and governable as you scale from 500 to 100,000+ users.

What’s new since the original guide

• Limits and guardrails: Atlassian now publishes customer-facing limits and recommended guardrails that correlate strongly with user experience and performance. Think of limits as enforced ceilings and guardrails as best-practice thresholds. When you approach or exceed them, expect slower pages, heavier admin overhead, and blocked scale efforts until you streamline.

• Site Optimizer (Enterprise & premium): A central hub that shows your instance health against thresholds, pinpoints what’s unused or oversized (e.g., custom fields, spaces, role actors), and lets you bulk clean or archive—with restore options for confidence. It reduces detective work and accelerates cleanups that used to take weeks.

• Custom templates (Enterprise): Standardize your best ways of working as reusable templates for company-managed and team-managed spaces. Codify field sets, workflows, permissions, and naming. This curbs config sprawl at the source—one of the most upvoted asks in the original comments.

 

Additional entities to cleanup

💡 Pro Tip: We recommend backing up your instance and running any changes in a test environment first before applying them in production.

Adding to the excellent list on Jira Instance hygiene for entities to audit, the following entities should also be included in audits.

Audit item	Audit resource
Statuses	What is a workflow status and significance of changing them How to view statuses Select Settings ⚙️ from the global navigation → Work items → Statuses
Permissions	Manage permissions in Jira Cloud | Atlassian Support
Filters & JQL	Manage filters | Jira Cloud | Atlassian Support Ensure that filters provide a specific list of spaces the data is searched over if possible to improve performance. To follow performance best practice, always specify the list of spaces you wish to search for in JQL. This will ensure you end up with consistent performance as your site grows.
Resolutions	How to view resolutions Select Settings ⚙️ from the global navigation → Work items → Resolutions
Fields & Field configuration custom fields field context screens screen schemes work type screen schemes	How do I set up fields in my Jira site? | Atlassian Support Create or edit a field configuration | Atlassian Support Change which field configuration scheme a space uses | Atlassian Support Manage work item screens | Atlassian Support Configure field contexts in your site | Atlassian Support

---

Enterprise hygiene pillars

1) Design for scale: control data shape

• Control custom fields: Consolidate duplicates, prefer shared contexts, and retire field options that create wide index footprints. Use Site Optimizer’s “last used” and recommendations to safely delete or archive fields and options.

• Space portfolio hygiene: Archive inactive Spaces and work items. Keep space lists navigable; most performance complaints in large sites correlate with oversized, inactive, or fragmented space portfolios.

• Permission and role actor simplification: Replace user-level grants with group-based permissions, prune deactivated users from roles, and consolidate duplicated schemes. This reduces permission calc overhead and errors. As you scale, and are looking to delegate permissions to project admins, invest in migrating permission schemes to role based permission schemes over direct grants.

2) Prevent sprawl with custom templates

• Codify your “golden paths” into templates: default work types, statuses, transitions, screens, fields, components, and permission patterns. Templates lower the need for ad-hoc cloning and eliminate small differences that accumulate into scale blockers.

• Starter set by domain: engineering delivery, service intake, business ops, vendor onboarding. Map each to reporting needs to keep field vocabularies consistent.

• Guardrails at creation time: Use templates to stay under field-per-space and work-type-per-space ceilings, and to limit workflow/state proliferation.

• An advanced usage of custom templates is to codify your template into a repository and use the public create space api (The Jira Cloud platform REST API ) to create Jira space with a deep customisations (Enterprise only)

ℹ️ For sites that are not enterprise, you can get an approximate result by using the "Share settings with an existing space checkbox” Create a new space | Jira Cloud | Atlassian Support . While this is not as complete of an experience as the full feature custom templates experience, it does help control site sprawl. Alternatively you can use the Jira API to provide you more control over which scheme is used The Jira Cloud platform REST API . For advanced users, it’s also possible to create a form or Jira service desk that allows end users to choose which compatible combinations (as defined by the admin) they can choose and then use Jira automations to trigger this API call.

3) Observe and act continuously

• Treat hygiene as an operating rhythm, not a one-off space. Make it measurable and repeatable.

• Quarterly reviews: Check Site Optimizer trends for custom fields, spaces, role actors, and archived items. Track “time-to-green” from yellow/red states.

• Change governance: Require approvals for new global objects (fields, statuses, schemes). Reuse before create. Tie requests back to a template or an approved exception.

• We recommend that you create a space for this specific activity with a template Epic as a checklist for all of the auditing work required. You can then set up an automation rule to clone this sample epic (Automation rule to clone an Epic with its child issues | Jira and Jira Service Management | Atlassian Support ) every quarter.

---

 

From guardrails to action: what to monitor

ℹ️ Think in three layers: inventory size, shape complexity, and usage recency.

Entity (examples)	Why it matters	Signals to watch
Custom fields and options	Large field catalogs increase index size and page compute; options amplify cardinality.	Field count vs. thresholds “Last used” trending stale Duplicate semantics across teams
Spaces and work items	Overgrown portfolios slow navigation and inflate permission calculations and reindex scope.	Inactive spaces > 12 months Very large spaces without archival Aging epics/work items with no updates
Roles, permissions, user grants	User-level grants and zombie users bloat ACLs; group-based permissions scale better.	Role actor counts climbing Deactivated users still in roles Permission scheme sprawl

Using Site Optimizer effectively

For instructions, see What is the Site optimizer?. Use it when your instance trends towards guardrails, after migrations, and as part of quarterly reviews.

• Prioritize changes that are low-risk and high-impact: unused custom fields, deactivated role actors, and inactive spaces. Validate with ownership and last-used data before action.

• Prefer archive-then-delete to reduce risk and keep a rollback window. Communicate planned removals and provide field mappings to protect dashboards and integrations.

💡 Enterprise tip: Add Site Optimizer outcomes to your quarterly governance pack. Track: fields removed, spaces archived, role actors reduced, and user-perceived performance improvements.

Standards at creation time with custom templates

Custom templates let admins define authoritative configurations and apply them consistently across teams and sites. They address a top comment theme from the original post—“How do we keep things clean after we finish the cleanup?”

• Field catalogs: Only include approved fields with shared contexts. Link to data dictionary pages to make intent obvious.

• Workflow kits: Provide standardized status sets, resolution rules, and validators to avoid local variations that explode complexity.

• Permission posture: Prefer role based permissions. Bake in audit-friendly patterns to reduce future cleanup. 

💡Common enterprise flow: Start with “shallow” templates for company-managed spaces and expand with deeper templates only when governance maturity is high.

 

---

Answering popular questions

Q. How often should we clean up, and who should be involved?

A. Quarterly light cleanups and an annual deep-dive work well. Keep the decision group small (maximum 2–5 admins), consult space admins and owners for specific retirements, and document decisions in Confluence. Build a change calendar so archivals don’t surprise teams.

 

Q. What should we delete vs. archive?

A.

• Archive when there’s a plausible compliance or discovery need (old spaces, long-closed work items) or when you need a rollback path.

• Delete when it’s clearly redundant or unused (duplicate custom fields, deprecated role actors), and you’ve validated no critical reports depend on it.
   

Q. How do limits affect day-to-day work?

A. When you’re over a limit (e.g., fields per spaces), new additions are blocked until you streamline. Guardrails warn you before enforced ceilings cause friction. Use Site Optimizer to find quick wins that unblock teams without risky rework. 

 

Q. Team-managed vs. company-managed spaces and hygiene

A. Team-managed spaces are flexible but can fragment standards; company-managed spaces centralize controls. For enterprises, prefer company-managed with templates for net-new teams, and set criteria for when a team-managed space is acceptable. If consolidating, be mindful that some fields and estimations differ—plan for mapping.

 

Q. How do we keep performance high as we approach 100k users?

A. 

• Favor shared, minimal configurations; avoid per-team variants without a strong reason.

• Archive aggressively: inactive spaces and stale work items add real overhead at scale.

• Automate reviews: scheduled exports from Site Optimizer, governance dashboards, and alerts when attributes trend towards thresholds. 

A pragmatic, size-based playbook

• 500–1,000 users: Establish templates and a naming convention now. Run quarterly Site Optimizer reviews for fields and roles. Archive spaces inactive for 9–12 months.

• 1,000–10,000 users: Introduce an intake process for new global objects; require template alignment. Consolidate permission schemes, convert user grants to group grants, and enforce single-source field dictionaries.

• 10,000–100,000 users: Treat hygiene as SRE work. Set OKRs for staying in green on guardrails, monitor Site Optimizer weekly, and run rolling space archival. Consider federating work across multiple sites when business domains are truly independent, with common templates to keep reporting coherent.

Governance mechanics that stick

• Roles and RACI: Central Jira Admins approve global objects; Space Admins request and justify; a small review board adjudicates exceptions.

• Documentation: Keep a living Confluence “Jira Data Dictionary” and “Workflow Catalog.” Link these into custom templates and intake forms.

• Auditability: Use Site Optimizer bulk-action logs and change records in Confluence to demonstrate control for internal audit and regulators.

Putting it into motion: a 90-day plan

1. Baseline: Run Site Optimizer. Capture custom fields, options, Spaces (inactive), and role actors. Identify top 10% worst offenders across each.

2. Stabilize: Bulk-archive unused fields and inactive spaces; remove deactivated users from roles; convert user grants to groups. Draft or refresh a minimal field dictionary.

3. Standardize: Publish 3–5 custom templates. Wire an intake form for new global objects that forces reuse consideration. Schedule quarterly reviews and set alerting on threshold drift.

FAQ (enterprise-focused)

Q. Will Site Optimizer changes break reports or dashboards?

A. Use “last used” and space/screen context to validate safety. Communicate early: announce planned removals and provide field mappings. Prefer archive-then-delete to create a rollback window if you discover dependencies.

 

Q. How do we align templates with marketplace apps?

A. Start with a core, app-agnostic template. For critical apps (e.g., test management), publish an “app-compatible” variant that preserves required fields and events. Revisit app usage quarterly; remove app-specific fields when the app is retired.

 

Q. What if we’re already over limits?

A. You can still operate, but additions may be blocked. Your fastest route back under is: remove unused fields, consolidate work types, archive inactive spaces, and simplify permission actors. Site Optimizer is designed to surface and automate those steps.

 

✅ Key takeaway: Pair proactive standards (custom templates) with continuous observation and bulk cleanup (Site Optimizer). That combination keeps you within limits, protects performance, and scales governance from 500 to 100,000+ users.

 

---

Done a cleanup before? Already have a process for keeping your instance clean?
Let us know what advice you have for others doing this for the first time.👇

Darrel Jackson

Great update. I like it.

It would be useful for the original article to be updated to include these deltas so we have all the information in one place rather than having to cross reference two pages.