Am 13.02.2021 um 07:39 schrieb Frank Schilder <frans(a)dtu.dk>dk>: The graphs were forgotten or filtered out. Changing the buffered_io value on one host will not change client IO performance as its always the slowest OSD thats decisive. However, it should have an effect on the IOP/s load reported by iostat on the disks on the host. Does setting bluefs_buffered_io=true on all hosts have an effect on client IO? Note that it might need a restart even if the documentation says otherwise. Best regards, ================= Frank Schilder AIT Risø Campus Bygning 109, rum S14 ________________________________________ From: Björn Dolkemeier <b.dolkemeier(a)dbap.de> Sent: 13 February 2021 07:16:06 To: ceph-users(a)ceph.io Subject: [ceph-users] Latency increase after upgrade 14.2.8 to 14.2.16 Hi, after upgrading Ceph from 14.2.8 to 14.2.16 we experienced increased latencies. There were no changes in hardware, configuration, workload or networking, just a rolling-update via ceph-ansible on running production cluster. The cluster consists of 16 OSDs (all SSD) over 4 Nodes. The VMs served via RBD from this cluster currently suffer on i/o wait cpu. These are some latencies that are increased after the update: - op_r_latency - op_w_latency - kv_final_lat - state_kv_commiting_lat - submit_lat - subop_w_latency Do these latencies point to KV/RocksDB? These are some latencies which are NOT increased after the update: - kv_sync_lat - kv_flush_lat - kv_commit_lat I attached one graph showing the massive increase after the update. I tried setting bluefs_buffered_io=true (as it’s default value was changed and it was mentioned as performance relevant) for all OSDs in one host but this does not make a difference. The ceph.conf is fairly simple: [global] cluster network = xxx fsid = xxx mon host = xxx public network = xxx [osd] osd memory target = 10141014425 Any ideas what to try? Help appreciated. Björn -- dbap GmbH phone +49 251 609979-0 / fax +49 251 609979-99 Heinr.-von-Kleist-Str. 47, 48161 Muenster, Germany http://www.dbap.de dbap GmbH, Sitz: Muenster HRB 5891, Amtsgericht Muenster Geschaeftsfuehrer: Bjoern Dolkemeier _______________________________________________ ceph-users mailing list -- ceph-users(a)ceph.io To unsubscribe send an email to ceph-users-leave(a)ceph.io

Am 13.02.2021 um 09:23 schrieb Frank Schilder &lt;frans(a)dtu.dk&gt;dk>: Ahh, OK. I'm not sure if it has that effect. What people observed was, that rocks-DB access became faster due to system buffer cache hits. This has an indirect influence on data access latency. The typical case is "high IOPs on WAL/DB device after upgrade" and setting bluefs_buffered_io=true got this back to normal also improving client performance as a result. Your latency graphs look actually suspiciously like it should work for you. Are you sure the OSD is using the value? I had problems with setting some parameters, I needed to include them in the ceph.conf file and restart to force them through. A sign that bluefs_buffered_io=true is applied is rapidly increasing system buffer usage reported by top or free. If the values reported are similar for all hosts, bluefs_buffered_io is still disabled. If I may ask, what framework are you using to pull these graphs? Is there a graphana dashboard one can download somewhere or is it something you implemented yourself? I plan to enable prometheus on our cluster, but don't know about a good data sink providing a pre-defined dashboard. Best regards, ================= Frank Schilder AIT Risø Campus Bygning 109, rum S14 ________________________________________ From: Björn Dolkemeier &lt;b.dolkemeier(a)dbap.de&gt; Sent: 13 February 2021 08:51:11 To: Frank Schilder Cc: ceph-users(a)ceph.io Subject: Re: [ceph-users] Latency increase after upgrade 14.2.8 to 14.2.16 Thanks for the quick reply, Frank. Sorry, the graphs/attachment where filtered. Here is an example of one latency: https://drive.google.com/file/d/1qSWmSmZ6JXVweepcoY13ofhfWXrBi2uZ/view?usp=… I’m aware that the overall performance depends on the slowest OSD. What I expect is that bluefs_buffered_io=true set on one OSD reflects in dropped latencies for that particular OSD. Best regards, Björn Am 13.02.2021 um 07:39 schrieb Frank Schilder <frans@dtu.dk<mailto:frans@dtu.dk>>: The graphs were forgotten or filtered out. Changing the buffered_io value on one host will not change client IO performance as its always the slowest OSD thats decisive. However, it should have an effect on the IOP/s load reported by iostat on the disks on the host. Does setting bluefs_buffered_io=true on all hosts have an effect on client IO? Note that it might need a restart even if the documentation says otherwise. Best regards, ================= Frank Schilder AIT Risø Campus Bygning 109, rum S14 ________________________________________ From: Björn Dolkemeier <b.dolkemeier@dbap.de<mailto:b.dolkemeier@dbap.de>> Sent: 13 February 2021 07:16:06 To: ceph-users@ceph.io<mailto:ceph-users@ceph.io> Subject: [ceph-users] Latency increase after upgrade 14.2.8 to 14.2.16 Hi, after upgrading Ceph from 14.2.8 to 14.2.16 we experienced increased latencies. There were no changes in hardware, configuration, workload or networking, just a rolling-update via ceph-ansible on running production cluster. The cluster consists of 16 OSDs (all SSD) over 4 Nodes. The VMs served via RBD from this cluster currently suffer on i/o wait cpu. These are some latencies that are increased after the update: - op_r_latency - op_w_latency - kv_final_lat - state_kv_commiting_lat - submit_lat - subop_w_latency Do these latencies point to KV/RocksDB? These are some latencies which are NOT increased after the update: - kv_sync_lat - kv_flush_lat - kv_commit_lat I attached one graph showing the massive increase after the update. I tried setting bluefs_buffered_io=true (as it’s default value was changed and it was mentioned as performance relevant) for all OSDs in one host but this does not make a difference. The ceph.conf is fairly simple: [global] cluster network = xxx fsid = xxx mon host = xxx public network = xxx [osd] osd memory target = 10141014425 Any ideas what to try? Help appreciated. Björn -- dbap GmbH phone +49 251 609979-0 / fax +49 251 609979-99 Heinr.-von-Kleist-Str. 47, 48161 Muenster, Germany http://www.dbap.de dbap GmbH, Sitz: Muenster HRB 5891, Amtsgericht Muenster Geschaeftsfuehrer: Bjoern Dolkemeier _______________________________________________ ceph-users mailing list -- ceph-users(a)ceph.io To unsubscribe send an email to ceph-users-leave(a)ceph.io

Am 13.02.2021 um 09:42 schrieb Frank Schilder &lt;frans(a)dtu.dk&gt;dk>: For comparison, a memory usage graph of a freshly deployed host with buffered_io=true: https://imgur.com/a/KUC2pio . Note the very rapid increase of buffer usage. OK, so you are using a self-made dashboard definition. I was hoping that people published something, I try to avoid starting from scratch. Best regards and good luck, ================= Frank Schilder AIT Risø Campus Bygning 109, rum S14 ________________________________________ From: Björn Dolkemeier &lt;b.dolkemeier(a)dbap.de&gt; Sent: 13 February 2021 09:33:12 To: Frank Schilder Cc: ceph-users(a)ceph.io Subject: Re: [ceph-users] Latency increase after upgrade 14.2.8 to 14.2.16 I will definitely follow your steps and apply bluefs_buffered_io=true via ceph.conf and restart. My first try was to update these dynamically. I’ll report when it’s done. We monitor our clusters via Telegraf (Ceph input Plugin) and InfluxDB and a custom Grafana dashboard fitted for our needs. Björn