The storage controller is the compute part of a storage array that runs the storage software. If the storage controller fails, the storage software can’t run and the array goes offline. As a result, redundant storage controllers are critical to maintaining data accessibility. However, storage vendors have differing opinions on what is the best way to deliver high availability.
Active-Passive Controllers
Active-Passive storage controllers are by far the most common method of delivering high availability. In this configuration, two controllers are installed in the storage array but the LUNs are assigned to one controller. Although the second controller has access to the volumes/LUNs, it essentially sits idle waiting for the first controller to fail. If and when the first storage controller fails, the second controller steps in so that access to data continues.
Active-Active Controllers
An alternative to active-passive controllers is active-active controllers. With the active-active approach, both controllers are available to support LUNs and volumes. Both storage controllers can have LUNs assigned to them. If a controller fails, the surviving controller can support its and the failed controller’s LUNs.
Active-Active vs. Active-Passive
Deciding between the two approaches comes down to prioritization. The organization needs to decide what is more important; maximum performance while everything is working or predictable performance if there is a failure. The advantage of an active-active design is that when everything is working, there is more available storage processing power. This means the system can deliver better performance or support more workloads. During a failure, 100 percent of the I/O has to be handled by the surviving controller until IT replaces the failed controller. If both controllers were running at high utilization rates, then users and applications may experience a drop in performance.
With an active-passive design, if a controller fails the same processing power becomes available from the standby controller and there is no drop in performance. The downside to an active-passive design is that a “failed state” is not normal, meaning the passive controller is idle for most of its life. As a result, that compute goes to waste.
Deciding on a Controller Design
Most IT professionals will choose an active-active design when possible. They will risk a performance drop-off in the future for an increased level of performance now. Again, storage controller failure is rare so it is understandable why IT professionals would want to have the increased processing power available to them. When there is a failure, their users need to be prepared for a drop in performance.
The government of the County of Nevada’s IT Department considered active-active controllers an important requirement as they went through their storage system selection process. Join our on demand webinar as we interview them about their recent storage selection. Learn why they chose a Hybrid Array over an All-Flash array and how their environment has improved since then.
Hi George I don’t know of any block storage products still using active/passive…. what company are you referring to?
Dell EqualLogic is using Active/Passive (Active Stand By seems more accurate).
Like Active-Active refer to more than one design (Symmetric, ALUA)
Regards
Eric Diologeant
XtremIOPS