Scheduled server backups. No scripts. No guessing.
No rclone config, no crontab, no wondering whether last night's job actually ran. Fill one form, test the connection before you save, and watch every byte move.
How it works
Step 01 - You
Point.
Name the job, pick the folder, choose a bucket. One form, no rclone config.
Step 02 - Cron
Schedules.
Manual, an interval, or your own cron. CtrlOps writes the crontab for you.
Step 03 - It
Proves.
Live speed, files and ETA while it runs - then the full log of every run.
One of these ends with hope. The other ends with proof.
Both sides are setting up the same nightly copy of /var/www/html to S3. Left: install rclone, sit through the config prompts, write the cron line, and then find out nothing tells you whether it ran. Right: fill the form, test the connection, done. Watch which one still has a question at the end.
This is the Backup tab. Go on, drive it.
A working copy, wired to a real little model of a server. Install the engine, fill the form, test the connection, watch a job run. Then point one at a path that does not exist, and watch it succeed anyway.
Rclone Not Installed. It is the engine that moves your files. Nothing backs up until it is on the server.
Install rclone to get started.
Nothing here leaves your browser. In the app, files go straight from your server to your bucket - CtrlOps copies them off the box and shows you it worked. It does not pull them back, and it does not know what a database is.
One extra w in the source path. Six weeks of backups, all green, all empty.
Two green badges. One of them copied nothing.
Neither badge is lying. rclone really did succeed both times - the second job just had nothing to copy, because the source path has a typo in it. A run that transfers zero files is still a successful run. That is not a bug, it is what exit 0 means.
A cron line would have told you nothing at all. No output, no status, no record - just a job that quietly does nothing every night until the day you need it. This is the single most common way a backup turns out not to exist.
CtrlOps will not page you about it - you still have to look. But the file count is on the row, the log is one click behind it, and Test Connection catches the other half before you ever save. That is the difference between visible in two clicks and invisible for six weeks.
Every field that matters, and what to put in it.
This is the rclone config you never have to write, laid out as a form. Six providers, a storage class for the stuff you hope to never open again, and three knobs for when the upload is slower than the wire.
Six S3-compatible providers
AWS S3, Cloudflare R2, Backblaze B2, Wasabi, MinIO, and DigitalOcean Spaces. Pick yours from the dropdown instead of remembering its endpoint URL. The ones that need a Custom Endpoint have a field for it - leave it empty for AWS.
Glacier and Deep Archive
Storage Class is optional, and it is the field that decides your bill. Leave it on Standard when you want the files back quickly. Pick Glacier or Deep Archive for the copies you keep for a year and hope to never open.
Tune the transfer
You can skip all three - the defaults are fine for most setups. When an upload is slower than the wire should allow, check Bandwidth Limit first, then raise Transfers.
Or store no keys at all
Access and secret keys are stored encrypted. If the server already has IAM permissions for the bucket, toggle Use IAM Role on and skip both fields - then there is nothing to rotate and nothing to leak.
Rclone said no. Why?
Paste the failing line out of View Log into the AI Terminal and ask what it means. Permission and install failures are exactly what the docs send you there for.
The feature you only notice when it is missing.
Nobody writes a paragraph about a job that runs at 2 AM and says nothing. One person brought backups up unprompted, so one person is what you get here - a wall of quotes about this would be the suspicious version.
In your bucket. Not in ours.
Most backup services answer this question by taking custody of your data and charging you rent on it. CtrlOps never touches your files. It configures the copy, starts it, and shows you it worked.
CtrlOps is not in the path
Your files go from your server to your bucket. rclone does the transfer, over your own credentials, directly. Nothing routes through us, so there is no copy of your data on our side to leak, subpoena, or lose.
Or store no keys at all
Access keys are stored encrypted. Better still, switch on Use IAM Role and there is nothing to store: the server already has permission to write to the bucket, and CtrlOps never sees a secret.
The bucket is yours
You pick the provider, the region, and the storage class. You get the bill, at their prices, with Glacier and Deep Archive on the table. Cancel CtrlOps tomorrow and every backup you have is still sitting where you put it.
What this is not
This gets a copy of your files off the server, on a schedule, and proves it. That is the whole job. "Backup" is a word people load with a lot of other expectations, so here are the edges:
- It does not restore.CtrlOps gets a copy off the server. Pulling it back is your provider's console, the AWS CLI, or rclone copy the other way. Test that path yourself once a month - a backup you have never restored is only a rumour.
- It does not rotate or prune.Sync mirrors the source, so deletions on the server become deletions in the bucket. Copy only ever adds, and grows forever. Ageing data out is a lifecycle rule in your bucket, not a setting here.
- It does not know what a database is.It copies files. A live /var/lib/mysql caught mid-write is not something you can restore from. Dump it to a folder first, then back the folder up.
- S3-compatible destinations only.Six providers today: AWS, Cloudflare R2, Backblaze B2, Wasabi, MinIO, DigitalOcean Spaces. No Dropbox, no Google Drive, no FTP. More destinations are on the way.
- It needs rclone, and cron for schedules.Both install in one click from the Backup tab, but they do have to be on the server. Manual runs work without cron.
If you need orchestrated restores, point-in-time recovery, or database-aware snapshots, buy something built for that - it is a real category and this is not trying to be in it. What CtrlOps is for is the thing most servers are actually missing: nothing is copied off the box at all, and nobody has checked in months. When a run does fail, take the log line to the AI Terminal, or check Infra Monitoring if the disk filled up underneath it.
More you can do with CtrlOps.
Real-Time Infra Monitoring
Live CPU, memory, disk, and network for every server, streamed as it happens.
AI Terminal
Describe what you need in plain English. Review and approve every command before it runs.
Script Directory
Save the commands you retype every week, with variables filled in at run time. One click on any server.
Questions before you need it.
Get a copy off the box. Tonight.
Point CtrlOps at a folder, test the connection, and let it run at 2 AM. No rclone config, no crontab. 1-month free trial, no credit card.
✓ Start instantly·✓ No credit card·✓ No sneaky autorenewals
