Using iperf, let's get a baseline for our clusters' network performance:
-
Install iperf on the ceph-node1 and ceph-node2:
# sudo yum install iperf -y
- On ceph-node1, execute iperf with -s for the server, and -p to listen on a specific port:
# iperf -s -p 6900
You can skip the -p option if the TPC port 5201 is open, or you can choose any other port that is open and not in use.
- On ceph-node2, execute iperf with the client option, -c:
# iperf -c ceph-node1 -p 6900
You can also use the -p option with the iperf command to determine the number of parallel stream connections to make with the server. It will return a realistic result if you have a channel-bonding technique such as LACP.
This shows that we have network connectivity of ~2.15 Gbits/s between the Ceph nodes. Similarly, you can perform a network bandwidth check for the other nodes of your Ceph cluster. The network bandwidth really depends on the network infrastructure you are using between your Ceph nodes.