How StratoFusion monitors the systems behind your cloud transfers

Moving a folder between cloud providers can take much longer than moving a small file. The work may include listing thousands of items, comparing folders, waiting through provider rate limits and confirming that each child operation reaches a useful outcome.

That means a cloud-transfer product needs to monitor more than whether its website loads.

StratoFusion now gives authorised administrators a single read-only view of the systems behind its production transfer workflows. The page is not a customer control panel, but the signals it brings together support a more dependable customer experience.

A representative busy period: multiple transfers are running, more are queued, and the underlying services remain healthy.

What is being observed?

The overview checks several independent parts of the production system.

The application and database

The application coordinates connected accounts, validates requests, records jobs and reports progress. Its readiness check confirms that the application can reach the database it depends on.

This matters because a page can be online while an essential dependency is not. Treating application and database readiness as one explicit check makes that difference visible to an operator.

The transfer worker

Large copy, move, backup and sync operations run through a dedicated transfer worker rather than depending on a browser tab staying open.

The dashboard checks whether that worker is healthy and shows its current capacity:

• operations running now;
• operations waiting in the queue;
• allowed concurrency;
• maximum queue size.

An operator can therefore distinguish an idle worker from a busy one and a busy worker from an unavailable one.

Host resources

The production runtime also reports CPU, memory, disk, uptime and network activity. These measurements help identify pressure that may affect application or transfer performance.

A number is shown only when it was actually observed. If monitoring data is missing or stale, the dashboard says so instead of substituting zero and making the system look idle.

Monitoring health

The monitoring system checks its own data sources. StratoFusion shows whether host metrics, transfer metrics and Prometheus monitoring targets are currently being collected.

This prevents an important false conclusion: no reported activity does not necessarily mean nothing is happening. It may mean a monitoring source is unavailable. Those states are presented differently.

Why separate checks are useful

A transfer workflow crosses several boundaries. The application may be healthy while a cloud provider is rate-limiting requests. The worker may be healthy while a monitoring target is temporarily unavailable. A database problem may affect job state even when the public page still responds.

One generic “system online” light would hide those distinctions.

Separate signals help an operator choose a more appropriate response. They also support clearer customer-facing outcomes: retry later, reconnect an account, review a partial result or contact support are different actions and should not all be presented as the same error.

Monitoring does not grant infrastructure control

The dashboard is intentionally read-only.

The StratoFusion web application does not receive access to a server shell, Docker controls or virtual-machine power operations in order to show these measurements. Monitoring data is collected through narrow server-side interfaces, and private monitoring addresses are not sent to the browser.

That boundary reduces the authority of the customer-facing application while still giving operators useful visibility.

Knowing which release is running

The overview also shows the exact software release currently deployed.

Each production application image is built from a specific Git revision and carries that identity with it. The release shown by the dashboard comes from the running artifact rather than a label someone must remember to update manually.

That helps connect an observed change with the code that introduced it and makes rollback or investigation more precise.

What this means for customers

Infrastructure monitoring cannot promise that every provider request will always succeed. Google Drive, OneDrive and Dropbox remain independent services with their own limits, authentication requirements and occasional failures.

What it can do is reduce ambiguity inside StratoFusion's part of the workflow.

Operators can see whether core services are healthy, whether transfer capacity is occupied, whether the host has resource pressure, whether monitoring coverage is complete and which release is active. Missing evidence is shown as missing rather than quietly treated as success.

That supports a simple product goal: when cloud file work is complicated behind the scenes, the interface and support experience should still be calm, specific and honest.