AOS 6.1 Improvements for Day-2 database operations.

AOS 6.1 greatly improved database performance on Nutanix especially when the guest VM uses just a single disk for all the database files. The underlying change is known as vdisk sharding. Basically it allows the Nutanix CVM to scale up the number of threads used to service a single virtual disk under heavy load.

IOPS performance increased by almost 2X under a TPC-C like transactional benchmark on Microsoft SQL Server running HammerDB.

Performance Improvement IO from TPC-C Like workload

We also found that AOS 6.1 improved throughput by around 30% when doing a disk based full restore from within SQL server. In this experiment all the SQL Datafiles are on Drive N:\ and the backup files are on Drive M:\

SQL Server Datafile layout

On my test cluster, the average throughput across the entire restore process was around 750 MB/s and the peak throughput was around 1200 MB/s.

Restore Process Average Throughput 755MB/s

It turns out that 1200MB/s is about the maximum per-node throughput achievable due to network bandwidth limitations in my lab environment.

Peak Write Throughput ~1200MB/s

We see that the writes are fully replicated across the cluster, the Prism chart below shows the network bandwidth used by the data replication traffice during the restore.

Replication traffic during the restore process

All in all a very nice performance boost purely from a software update.

Leave a Comment