Skip to main content

Performing basic infrastructure checks

Some basic diagnostic checks of your infrastructure will help you rule out obvious sources of performance problems. You should review protocol and network settings and check disc throughput and latency. If you are replicating data, you will want to monitor throughput and latency between nodes.

  • Checking protocol settings on the storage system
    You can check that a performance issue is not related to protocol settings on your storage system. If the settings are the issue, you can take corrective action and then verify that the performance issue is resolved.
  • Checking the network settings on the data switches
    You must maintain the same network settings on your clients, storage systems, and switches to ensure that performance is not impacted. All components in the network must have the same MTU setting for best performance.
  • Checking the MTU network setting on the storage system
    You can change the network settings on the storage system if they are not the same as on the client or data switches. Whereas the management network MTU setting is set to 1500, the data network MTU size should be 9000.
  • Checking disk throughput and latency
    You can check the disk throughput and latency metrics for cluster nodes to assist you in troubleshooting.
  • Checking throughput and latency between nodes
    You can use the network test-path command to identify network bottlenecks, or to prequalify network paths between nodes. You can run the command between intercluster nodes or intracluster nodes.