On 11/15/19 12:57 PM, Willi Schiegel wrote: > Hello All, > > I'm starting to setup a Ceph cluster and am confused about the > recommendations for the network setup. > > In the Mimic manual I can read > > "We recommend running a Ceph Storage Cluster with two networks: a public > (front-side) network and a cluster (back-side) network." > > In the Nautilus manual there is > > "Ceph functions just fine with a public network only, but you may see > significant performance improvement with a second “cluster” network in a > large cluster. > > It is possible to run a Ceph Storage Cluster with two networks: a public > (front-side) network and a cluster (back-side) network. However, this > approach complicates network configuration (both hardware and software) > and does not usually have a significant impact on overall performance. > For this reason, we generally recommend that dual-NIC systems either be > configured with two IPs on the same network, or bonded." > > Am I misunderstanding something or is "significant performance > improvement" and "does not usually have a significant impact on overall > performance" in the Nautilus doc contradictory? So, which way to go? >
There is no need to have a public and cluster network with Ceph. Working as a Ceph consultant I've deployed multi-PB Ceph clusters with a single public network without any problems. Each node has a single IP-address, nothing more, nothing less. The whole idea of a separated public/cluster dates back from the time when 10G was expensive. But nowadays having 2x25G per node isn't that expensive anymore and is sufficient for allmost all use cases. I'd save the money for a second network and spend it on a additional machine in the cluster. That let's you scale out even more. My philosophy: One node, one IP. I've deployed dozens of clusters this way and they all work fine :-) Wido > Thank you very much > Willi > > > _______________________________________________ > ceph-users mailing list > ceph-users@lists.ceph.com > http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com _______________________________________________ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
