The following are example scenarios that may affect your Crossware services:
- Changes to connectors or mail flow rule have not taken affect (any change can take up to 1 hour)
- There is an issue with Microsoft connectivity between M365 and Azure (due to MS changes or networking issues in their data centres)
- The Crossware service has stopped working. Very unlikely because of our 3 data centre / multi-server redundancy scheme
Here is the process and how redundancy is handled:
The movement of the email from your M365 server to the Crossware Azure service and back again to your M365 server is managed by:
1. Crossware Mail Flow Rule
2. Crossware Outbound Connector
3. and the Crossware Inbound Connector
2. Crossware Outbound Connector
3. and the Crossware Inbound Connector
If the Connector cannot connect to Crossware Azure service, the email will fail (for test emails) and for normal emails, the emails will sit in your M365 SMTP queue until the email can be processed or the email times out and is returned to the sender as a "delivery failure"
An email is not accepted by the Crossware service unless, it can verify that the email can be immediately handed back to your M365 server via the Inbound connector. If it cannot, then the email will stay in your M365 SMTP queue until the email can be processed or the email times out and is returned to the sender as a "delivery failure"
If at any time you need to disable the Crossware service for mail flow issues: Delete these three documents. Or for temporary disablement, disable the Mail Flow Rule. Any of these changes can take up to 1 hour to take affect in the M365 cloud.
Redundancy:
Your Crossware service will have redundancy in this order depending on where you are based for this example we will use US:
1. MS Azure - US
2. MS Azure - Europe
3. MS Azure - APAC
2. MS Azure - Europe
3. MS Azure - APAC
If the US Azure cloud becomes unavailable, your emails will be processed by the next Azure data center (as shown above). Each Crossware instance in these Azure data centres has multiple servers for load balancing.