Should the collector run on Sopht stack or on your stack?

By default, we advise to run it on Sopht stack. That way we are in charge of the monitoring, alerting, we can update the collectors more easily…

The client can choose to run it on its own stack. The main reasons are:

• if the software can only be accessed via a restricted Network (local, private, though IP filtering…);
• if there are confidential data that the client want to anonymize before they reach Sopht’s stack;
• if the client’s prefer to handle secrets on their side, without communicating them to Sopht.

The consequences are:

• the customer handles the infra: container deployment, monitoring, debugging;
• the customer needs to update the collector to new versions when asked by Sopht, for instance for security or functional updates.

Retrieval on Sopht Stack

Retrieval on customer’s side

<aside> 💡

Azure Local

If you have Azure Local clusters, we will automatically fetch node metrics (used CPU, RAM and storage per node).

No special configuration is required on your side

</aside>