While the capacity varies with usage, the journal reservations (specifying the maximum capacity we will allow the replication processes to consume) must be set manually. Appropriately sizing journal volume reservations is critical to the health of the PowerFlex cluster, especially during WAN outages and other failure scenarios. For example, the journal volume must have enough available capacity to continue ingesting replication data even when the SDR cannot ship the journal intervals to the target site. If the journal intervals are unable to transmit, the journal buffer capacity will increase, potentially filling it altogether. So, you must consider the maximum cumulative writes that might occur in an outage. If the journal buffer space fills completely, the replica-pair volumes will require reinitialization.

The administrator sets and adjusts the maximum reservation size of the journal volumes. The minimum requirement for journal capacity is 28 GB multiplied by the number of SDR sessions, where SDR sessions equal the number of SDRs installed plus one. However, some additional calculation is required. The reservation size is stated in the system as a percentage of the storage pool in which each journal volume is contained. As a rule, reserve at least 5 percent of the storage pools for replication journals.

The reserved capacity for journals may be split into several volumes across multiple storage pools, or the replication journals may all reside in one storage pool of a protection domain. The performance character of any storage pool in which a journal volume resides must match or exceed the performance requirements of any storage pool in which the replicated volumes reside.

The journal capacity must be sufficient to accommodate factors such as volume overhead, free space reservations (to sustain node failures or accommodate Protected Maintenance Mode). The single most important consideration in sizing the journal capacity is a possible WAN outage. If we account for this scenario, we end up accounting for all the other considerations.

To begin, assess the journal capacity needed per application. We need to know the maximum application write bandwidth during the busiest hour, because application I/O varies over time, and we cannot predict when an outage might occur. The minimum outage allowance is 1 hour, but we strongly recommend using three hours in the calculations.

Calculation example:

• Our application generates 1 GB of writes during peak hours.
• Using 3 hours as the supported outage, we calculate from 10,800 seconds.
• The journal capacity reservation needed is 1 GB/s * 10800 s = ~10.547 TB.
• Because journal capacity is calculated as a percentage of storage pool capacity, we divide the needed space by the storage pool usable capacity. Let us assume that usable capacity is 200 TB.
• 100 * 10.547 TB / 200 TB = 5.27%.

As a safety margin, we will round up to 6%.

Repeat the calculation for each application being replicated.

Note: As the size and capacity of a storage pool changes, the percentage will change. Readjustments to the journal reservation will be necessary as pool capacities vary. Administrators can adjust the journal capacity reservation percentage at any time from the UI, CLI, or API.