SAS Resiliency

When adding Drive Enclosures to the second and third location within a string, you must disconnect one of the two cables connecting the string to the Controllers. To support continuous access to the data on the string during the time of this disconnection, the Oracle FS System reroutes all traffic to that string to the remaining Controller connection.

It can take the system up to 30 seconds for the system to reroute the data traffic. During this time, the system is vulnerable to additional failures or errors. To verify whether the system has stabilized, log into Oracle FS System Manager (GUI) and monitor the System Alerts page (System > Alerts and Events > System Alerts).

Expect event messaging that shows the topology rule violations that corollate exactly to the cable pulls. If, for example, you disconnect a Drive Enclosure port in the path of Controller 01, you should expect messaging related to Controller 01.

You might see up to 48 of the same message, one per drive with one disconnected cable. The following table shows examples of messaging to expect.

Table 1 Expected messaging
Types of messages	Example of messaging to expect
Alerts	`Fabric Topology Warning (on System Alert Screen)` `PITMAN Topology Change Warning (on Event log)` `Foreign Drive Type (on System Alert Screen)`
Warnings	`Enclosure Component State Change (on Event log)` `Storage Fabric Path Failure (on Event log)`
Informational	`Degraded Connectivity (on Event log)` `PITMAN Topology Change OK (on Event log)` `New Drive Group (on Event log)`

If you receive events other than the expected events, correct them prior to proceeding to ensure there are no unexpected or compromising issues. After the system stops generating the expected events, you can proceed.