Cloud sync and working offline
Phloem Computer records on your computer first and uploads after. A lost network never stops a capture.
What needs the cloud
| Works | Needs |
|---|---|
| Capture and recording | Nothing. Works offline and without signing in. |
| Deploy a Trellis setup, Start run | A computer that has signed in once. Then it works offline. |
| Cloud sync, remote replay | Signed in |
| Hamming, Publish, Load published…, Recover from backup…, installing a connector from the catalog | Signed in and online |
A button that needs the cloud shows a cloud mark. Pressing it while signed out signs you in first. While offline, it is disabled and says why.
To use Computer with no account, press Use this computer locally on the first screen. Sign in later from Settings › Cloud connection › Sign in to cloud.
How captures catch up
Every capture is written to disk first. When the cloud is reachable, Phloem uploads it. When it is not, data waits on disk and uploads once the connection returns.
Captures wait within the Disk budget. When the budget is full, the oldest data not yet uploaded is deleted. An input over its share loses its own oldest data first.
Set the budget in Settings › Storage:
| Setting | Does |
|---|---|
| Disk budget | Space for recordings waiting to sync |
| Drain interval | How often buffered data is uploaded |
| Disk space warnings | Notifies you when runs fill the budget, or it will fill within the hour |
Disk budget by input shows how much each input holds.
Choose what to send
Open Settings › Cloud sync › Send to the cloud:
| Choice | Sends |
|---|---|
| Everything | All data, as it is captured |
| Runs only | Data inside a run, while the run is open. Data between runs stays on this computer. |
| Local only | Nothing. No data, and no project setup. The console still lists runs, marked "on device". |
When you widen the choice, Phloem sends what it kept before. The page says so while it runs.
To keep one input off the cloud, click the cloud icon on its row and press Keep local. Sync to cloud turns it back on.
Check what reached the cloud
| Where | Shows |
|---|---|
| The cloud icon on an input's row in Inputs | That input's state, and captured against in-cloud counts |
| Settings › Cloud sync | One line for the whole computer, then Upload, Time to catch up, Behind, Not in cloud and Disk free |
| Settings › Diagnostics | Missing from the cloud: each input's missing data, when and why. On this disk: damage and data set aside. |
phloem audit --last 1h | What was captured against what reached the cloud, for a window |
phloem doctor | Checks this computer's setup |
The cloud icon reads:
| Reads | Means |
|---|---|
| Up to date with the cloud | Everything captured has synced |
| Uploading · … queued | Syncing; the amount is still to send |
| Behind · … queued | A backlog is catching up. Nothing is lost. |
| Can't reach cloud · … queued | Kept here until the connection returns |
| Uploads failing · … queued | Uploads are failing; data is kept here |
| Not signed in — kept here | Sign in to sync |
| Local only — kept on this computer | Send to the cloud is Local only |
| Kept on this computer | Sync is off for this input |
While offline, Settings › Cloud sync shows Filling and Records for instead: how fast data builds up, and how long until the budget is full.
The budget can't use more than the disk's free space. When the disk is the limit, Disk budget says it is limited by disk. Check Disk free before a long offline run.
Updated