Move Whole Projects
A bulk move takes one project at a time and moves every file in it that passes Storage Saver's check, whoever uploaded it, so nobody has to pick files one by one. A Jira administrator starts it from the admin page's Move files tab; each start is a run, which the tab follows until it ends.
Before you start
- Wait for a finished scan. The tab lists projects from the last scan that finished, and a bulk move takes only files uploaded before that scan started; to include newer ones, press Scan again on the Overview tab first. While a scan is running, no bulk move can start.
- Show the panel on every work item in the project (see Work Item Panel). After the run, the files are gone from their work items' attachments, and people reach them through the Storage Saver panel, which Jira does not show until someone adds it.
- Set the minimum age if anything on your site reads recent attachments (see Settings). The run applies the value set when you confirm the move, and keeps it until the run ends.
- A bulk move does not use your own work item permissions: it moves the project's files whoever uploaded them, including files on work items you cannot see, and records you as the person who moved each one (see Permissions).
- In a service project, customers keep every file they see on the customer portal. The portal shows a request's files only where its description or a reply uses them, and Storage Saver never moves such a file.
What the tab lists
The table Projects with the most to move lists up to ten projects: those where the last scan found the most storage that could move, counting Can be moved now and Needs checking together. Other projects have no Move files… button in this report.
| Column | What it shows |
|---|---|
| Project | The project's name and key. |
| Can be moved now | Storage in files the scan proved unused in their work item's text. |
| Needs checking | Storage in files the scan could not place either way. The run checks each one and moves it only if it is not in use. |
| Not old enough yet | Only when the scan counted with a minimum age above 0: storage in files younger than it. The run does not move them. |
Each row ends with a Move files… button. When a button is off, the reason is under it: "Wait for the scan to finish before moving files.", "A bulk move is already running." or "The last scan found nothing here that a run could take."
Every scan builds the table afresh. To reach a project that is not listed, move the files of the projects above it and press Scan again, or pick the project's files on its project page, up to 25 at a time.

Move a project's files
Open the Move files tab
Go to Jira settings → Apps → Storage Saver and select the Move files tab.
Pick the project
Press Move files… in the project's row. Nothing moves yet; the button opens a dialog first.
Read the dialog and confirm
The dialog shows what the last scan found in the project, under Can be moved now and Needs checking, and the minimum age the run applies. Every file is checked again before it moves, so the result can differ from these figures. Under Carries on until the project is done it says that the move goes on with the page closed: "Files transfer one at a time, usually several hundred an hour and fewer when they are large." When an earlier run on this project ended before the project did, the dialog also says where this press starts (see Press again after a run).
Press Move files to start the run, or Cancel.

Nothing moves until you press Move files Follow the run
The run appears above the table. A line under its heading says what it is doing (see How a run moves a large project), and its counters grow as it reads the project's work items:
- Work items read: how far the run has got.
- Queued: files that passed the check and are on their way out of Jira.
- Refused by the check: files that stay in Jira, for example because they are used in a work item's text. A list below the counters names them with the reason.
- Left alone, newer than the scan: files uploaded after the scan started.
- Could not be queued: files the run could not hand over. A note says what to do.
The run carries on if you leave the page, until it has read the project's last work item.

A run in progress, with what it has queued and what the check refused Read the result
When the run ends, it moves below the table under Last bulk move, headed with the number of files it queued and their size, and stays there until the next run replaces it. The queued files go on moving one at a time after the run has ended. Each file's state shows on the project page and on its work item's Storage Saver panel.
How a run moves a large project
One press moves the whole project, however large. The run queues files in rounds: when a round is full, the run pauses, and a check that runs every hour continues it on its own, with no page open and no new press. A project with many thousands of files can take days.
While the run is live, the line under its heading reads one of these:
| The line reads | What it means |
|---|---|
| "Checking and queuing files." | The run is reading work items and queuing the files that pass the check. |
| "Waiting for about 600 queued files to transfer before queuing more." | Many of the run's files still wait to transfer, so it queues no more until they have moved, and looks again every few minutes. |
| "Slowed down: recent transfers ran out of retries, often because Jira is limiting downloads. It carries on more slowly." | Some files could not be transferred, usually because Jira was limiting downloads. The run queues nothing for 15 minutes, longer while such failures go on. |
| "Paused after queuing as much as one round can. It continues within the hour." | The round is full. The hourly check starts the next one. |
| "Paused because the job queue was unavailable. It tries again within the hour." | The run could not schedule its next step. The hourly check tries again. |
A run ends when it has read the project's last work item, when someone stops it (see Stop a run), or when something keeps it from going on; the result then says why. Jira limiting downloads only slows a run down: a run ends over failed transfers only when the files it queued keep failing, with none succeeding in between. The result then reads "Files this run queued kept failing to transfer, so it stopped queuing more. Files already queued will still be tried. The failures are listed under Needs attention."
A work item with more files than one round can queue is skipped and counted under Could not be queued; the run names it, and its files have to be moved from the work item itself.
Press again after a run
A run that ended before the project did, because it was stopped or something kept it from going on, keeps its place. Press Move files… in the same row again, and the dialog says where the new press starts:
- Continues the earlier run: "This press continues the run of", followed by its start, "it starts after the last work item that run decided." To start over instead, tick Start from the first work item instead.
- Starts again at the first work item: the minimum age is now lower than when the earlier run started, or a newer scan has finished and files the earlier run passed over, as too young or as newer than its scan, may now be old enough to move.
Each project keeps its own place, so a run on another project in between changes nothing. A newer scan keeps it too, as long as the minimum age still keeps out every file the earlier run passed over; with no minimum age set, a press after a newer scan starts again. Either way nothing is moved or counted twice: files the earlier run moved are no longer on their work items. After a run that read the whole project, a new press starts at the first work item.
Stop a run
Press Stop in the run's section. The dialog Stop this bulk move? says: "Files already queued will still be moved; nothing more will be queued." Press Stop to confirm.
- The run ends after the work item it is on; a run that is waiting or paused ends within seconds. Until then, the section says "A stop was requested by", followed by the name of the administrator who pressed it.
- Stopping never takes back a file that is already queued: it still moves. A moved file can be restored, one file at a time, from its work item's panel or the project page.
- Any Jira administrator can stop a run, not only the one who started it.
- The result then reads Stopped. Pressing Move files… again goes on as Press again after a run describes.
One at a time
Only one bulk move runs at a time, and files move one at a time across the whole site, moves and restores alike. A run can queue hundreds of files, so a move or restore started from a panel or a project page meanwhile waits its turn.
Export the project figures
Once a scan has finished, press Export projects as CSV beside the table's heading. It downloads the figures of the last finished scan for every project it read, not only the ten in the table: one row per project, sorted by project key. The tab confirms how many projects were exported, and that "The file lists figures only, never file names, and the download link works for an hour."
The columns of the exported file
| Columns | What they count |
|---|---|
project_key, project_name | The project. |
files, bytes | Every file the scan found in the project, and their size. |
can_be_moved_now_files, can_be_moved_now_bytes | Can be moved now. |
checked_when_moved_files, checked_when_moved_bytes | Needs checking. |
in_use_files, in_use_bytes | In use. |
not_old_enough_files, not_old_enough_bytes | Files younger than the minimum age the scan used. |
too_large_files, too_large_bytes | Files larger than 500 MB. |
Sizes are in bytes.
If a move does not finish
- A file refused by the check has not failed: it stays in Jira, and the run's Refused by the check list says why.
- A queued file whose move fails or gets stuck is handled like any other move: its tile on a panel or the project page reads Needs attention, and for a queued file that happens two hours after it was queued. While the run's files keep moving, the Saved files tab does not list each one still waiting its turn; it counts them in one line, and lists a file once it has waited 12 hours.
- If the last check before the delete cannot confirm that a file is unused, or Jira refuses the delete for good, Storage Saver discards its copy and the file stays in Jira as it was, with no entry in the app.