See what is in Object Storage
Goal
See which buckets your organisation has, how much is in each, look inside one, and remove files the platform has no record of.
Prerequisites
- You are signed in to the Admin Control Center as a platform administrator.
- Your installation runs a file store. Object Storage is optional; without one this screen shows no buckets and cannot create any.
Steps
- Choose Storage in the left-hand list, under Administration.
- The list shows every bucket your organisation has, with how many files are in each and how much space they take. A bucket is a named container — apps and integrations create their own, and you can make one here with New bucket.
- Select a bucket and choose Browse (or just click its row) to see what is inside.
Each file is listed by its key — the full path it is stored under, like
exports/2026/ledger.csv. - Type into Prefix to narrow the list to one folder, for example
exports/2026/. - Select a file and choose Download to get a copy, or Delete to remove it.
- Upload adds one or more files to the folder you are looking at: each file's own name is added to the prefix currently in the box. Pick several at once in the file dialog; they go up one after another, and the button counts them. If one fails the rest still go up, and a message names the first that did not — except when the storage limit is reached, which stops the rest.
- To go back, choose Back to buckets or use the trail at the top of the page.
What happens behind the scenes
This screen is not about one environment. Every other page under Administration works against the environment chosen in the top bar; this one does not. A bucket belongs to your whole organisation, and creating or deleting an environment neither creates nor removes anything here.
A record's own files are not shown. A photo or a document attached to a record belongs to that record, is served only to people allowed to read it, and is not listed here — even when the record is large enough that the platform kept it in the file store. Those files are reached from the record itself.
The counts come from the platform's own index, not from the store. That makes the page quick to draw, and it has one consequence worth knowing: a file uploaded straight to the store by a link that was never confirmed has no entry, so it is not counted and not listed. That is what Sweep is for.
How much you may store
The line under the title says how much your tenant stores, for example Using 12.4 MB of 100 MB. The figure is the files in your buckets and the files on your records together: a photo on a contact and a document in a bucket come out of the same allowance.
- A trial may store 100 MB. A paid plan sets the plan's included storage instead, or lifts the limit for a plan that states none; the line then reads No storage limit.
- When a file would go over the limit it is refused, with a message saying how much is used, how much is allowed and how big the file is. Nothing already stored is touched. Delete files you no longer need, or upgrade the plan.
- Replacing a file needs room only for the difference. Swapping a 5 MB photo for a 6 MB one needs 1 MB free, not 6.
Clearing out files nothing points at
Choose a bucket and select Sweep.
- Preview first. It reports how many files it examined, how many the platform has no record of, and how many were too recent to judge. Nothing is removed.
- Remove deletes what the preview found. It only becomes available once a preview has found something.
Why the preview matters. A sweep decides what to remove by what is missing rather than by what it put there itself, which is a riskier thing than an ordinary delete. Files written in the last day are always left alone, because an upload still in progress looks exactly the same as one nobody ever finished.
Troubleshooting
- "The bucket name is already taken" — bucket names are shared across every customer of
every Parsware installation, because that is how S3 storage works. A short name like
documentswill almost certainly be gone; use something specific to you. - The sweep refuses the bucket. It holds files belonging to records, which the platform keeps under a reserved folder. Those files are in use even though this screen has no record of them, so sweeping is refused rather than risking them. Sweep a bucket of your own instead.
- A file's hash column says "None". It was uploaded straight to the store rather than through the platform, so nothing ever read the bytes to fingerprint them. Uploading it through this screen gives it one.
- The size looks smaller than the disk. Same reason: unconfirmed uploads are not counted. A preview sweep will show them.
Related scenarios
- Create an environment.
- Add a file or picture to a record.