HPE Apollo 4500 Reference Manual page 14

Reference guide
Configuring disks
When array controller write cache is available, it is recommended to configure drives in RAID 0 with controller write cache enabled to improve
small object write performance.
For a fully disk-populated HPE Apollo 4510 Gen9 with 68 drives, significant CPU cycles must be reserved for 68 OSDs on a single compute node.
Configuring RAID 0 volumes across two drives at a time—resulting in 34 OSDs—could reduce CPU usage. Configuring multiple drives in a RAID
array can reduce CPU cost for colder storage in exchange for reduced storage efficiency to provide reliability. It can also provide more CPU
headroom for error handling or additional resources if cluster design dictates CPU resource usage outside of cluster specific tasks.
Choosing a network infrastructure
Consider the desired bandwidth of storage calculated in the preceding paragraph, the overhead of replication traffic, and the network
configuration of the object gateway's data network (number of ports/total bandwidth). Details of traffic segmentation, load balancer configuration,
VLAN setup, or other networking configuration/best practice are very use-case specific and outside the scope of this document.
• Typical choices of configuration for data traffic will be LACP bonded 10GbE links. These links provide resiliency if spanned across switches
and aggregated bandwidth.
• Network redundancy (active/passive configurations, redundant switching) is not recommended, as scale-out configurations gain significant
reliability from compute and disk node redundancy and proper failure domain configuration. Consider the network configuration (where the
switches and rack interconnects are) in the CRUSH map to define how replicas are distributed.
• A cluster network isolates replication traffic from the data network and provides a separate failure domain. Replication traffic is significant, as
there are multiple writes for replication on the cluster network for every actual I/O. It is recommended to bond all 10GbE links with LACP and
segment the public and backend traffic via VLANs.
• It is recommended to reserve a separate 1GbE network for management as it supports a different class and purpose of traffic than cluster I/O.
Matching object gateways to traffic
Start by selecting the typical object size and I/O pattern then compare to the sample reference configuration results. The object gateway limits
depend on the object traffic, so accurate scaling requires testing and characterization with load representative of the use case. Here are some
considerations when determining how many object gateways to select for the cluster:
• Object gateway operation processing tends to limit small object transfer. File system caching for GETs tends to have the biggest performance
impact at these small sizes.
• For larger object and cluster sizes, gateway network bandwidth is the typical limiting factor for performance.
• Load balancing does make sense at scale to improve latency, IOPS, and bandwidth. Consider at least three object gateways behind a load
balancer architecture.
• While very cold storage or environments with limited clients may only ever need a single gateway, two is the recommended minimum to
protect against a single point of failure.
With the monitor process having relatively lightweight resource requirements, the monitor can run on the same hardware used for an object
gateway. Performance and failure domain requirements dictate that not every monitor host is an object gateway, and vice versa.
Monitor count
Use a minimum of three monitors for a production setup. While it is possible to run with just one monitor, it's not recommended for an enterprise
deployment, as larger counts are important for quorum and redundancy. With multiple sites, it makes sense to extend the monitor count higher
to maintain a quorum with a site down.
Use physical boxes rather than virtual machines to have separate hardware for failure cases. It is recommended that the monitors utilize mirrored
SSDs due to the high number of fsync calls on these nodes.
Page 14