[ceph-users] Re: Increasing QD=1 performance (lowering latency)

2021-02-12 Thread Mark Nelson
FWIW, the current RDMA implementation is part of the async messenger: https://github.com/ceph/ceph/tree/master/src/msg/async/rdma Haomai probably is the leading authority on it, though we just got a shipment of new performance test gear for the community lab that theoretically will let us

[ceph-users] Re: Increasing QD=1 performance (lowering latency)

2021-02-12 Thread Max Krasilnikov
День добрий! Thu, Feb 11, 2021 at 04:00:31PM +0100, joachim.kraftmayer wrote: > Hi Wido, > > do you know what happened to mellanox's ceph rdma project of 2018? We tested ceph/rdma on Mellanox ConnectX-4 Lx during one year and saw no visible benefits. But it was strange connection outages

[ceph-users] Re: Increasing QD=1 performance (lowering latency)

2021-02-11 Thread Joachim Kraftmayer
Hi Wido, do you know what happened to mellanox's ceph rdma project of 2018? We will test ARM Ampere for all-flash this half-year and probably get the opportunity to experiment with software defined memory. Regards, Joachim ___ Clyso GmbH Am 08.02.2021 um

[ceph-users] Re: Increasing QD=1 performance (lowering latency)

2021-02-08 Thread Paul Emmerich
A few things that you can try on the network side to shave off microseconds: 1) 10G Base-T has quite some latency compared to fiber or DAC. I've measured 2 µs on Base-T vs. 0.3µs on fiber for one link in one direction, so that's 8µs you can save for a round-trip if it's client -> switch -> osd