<div dir="ltr">Pulp 3.14 is out, which includes both the content serving improvements and also "retry on error" downloads which should help the CDN errors you've been experiencing. Katello should have their RPMs ready in a couple of days.<br></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Mon, Jun 28, 2021 at 9:39 AM Daniel Alley <<a href="mailto:dalley@redhat.com">dalley@redhat.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr"><div>Sorry Bin, this ended up in my spam somehow so I missed your update until a second ago.<br></div><div><br></div><div>Realistically, it's probably getting bottlenecked on the database. You can definitely try increasing the workers further (beyond 50) but I'm not sure how much it will help. A lot of the improvements in 3.14 are oriented around reducing our load on the database so it should help quite a bit.<br></div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Tue, Jun 22, 2021 at 12:35 PM Bin Li (BLOOMBERG/ 120 PARK) <<a href="mailto:bli111@bloomberg.net" target="_blank">bli111@bloomberg.net</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div><div style="white-space:pre-wrap;font-size:small;font-family:"Courier New",Courier,"BB.FixedWidth"">We will look into upgrade 3.7.3 to 3.14. <div>For now, I have updated number of worker a few times. We are having 50 workers running. I no longer see the timed out messages but the TIME_WAIT is still around 5k.</div><div><br></div><div># netstat -an | grep -i TIME_WAIT |grep 24816 | wc -l</div><div>5473</div><div><br></div><div>Also notice the database connection is over 60.</div><div>=> select count(*) from pg_stat_activity where usename = 'pulp';</div><div> count </div><div>-------</div><div> 63</div><div>(1 row)</div><div><br></div><div>Should I keep adding workers until the queue comes down? We still have plenty of cpu and memory on the host.</div><div><br><br><div style="font-size:small;font-family:"Courier New",Courier,"BB.FixedWidth""><div><div><div>From: <a href="mailto:bmbouter@redhat.com" target="_blank">bmbouter@redhat.com</a> At: 06/22/21 12:01:30 UTC-4:00</div>To: <a href="mailto:danny.sauer@konghq.com" target="_blank"> danny.sauer@konghq.com</a><br>Cc: <a href="mailto:bli111@bloomberg.net" target="_blank"> Bin Li (BLOOMBERG/ 120 PARK ) </a>, <a href="mailto:pulp-list@redhat.com" target="_blank"> pulp-list@redhat.com</a><br>Subject: Re: [Pulp-list] Content server Performance</div><br></div><div style="background:white none repeat scroll 0% 0%;color:black;font-family:Arial,"BB.Proportional";font-size:small;white-space:normal"><div><blockquote><div dir="ltr"><div dir="ltr"><br></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Tue, Jun 22, 2021 at 11:56 AM Danny Sauer <<a href="mailto:danny.sauer@konghq.com" target="_blank">danny.sauer@konghq.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr">You can certainly run multiple instances of the content server. It just needs a connection to the database and access to the storage.</div></blockquote><div>Agreed, you could deploy additional content servers and have your nginx/apache load balance them.</div><div> <br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr"><div><br></div><div>Have you tuned the number of worker processes in Gunicorn? It defaults to 1, but should almost certainly be increased for any sort of volume. <a href="https://docs.gunicorn.org/en/stable/settings.html#worker-processes" target="_blank">https://docs.gunicorn.org/en/stable/settings.html#worker-processes</a></div></div></blockquote><div>Pulp changed the default gunicorn worker processes to 8 maybe a release or two ago. See the `pulp_content_workers` variable in the installer here <a href="https://pulp-installer.readthedocs.io/en/latest/roles/pulp_content/#role-variables" target="_blank">https://pulp-installer.readthedocs.io/en/latest/roles/pulp_content/#role-variables</a></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr"><div><br></div><div>There are several moving pieces, but that's really all I had to touch here.</div><div><br></div><div>--Danny</div></div></blockquote><div>With pulpcore==3.14 there is a significant performance improvement being reviewed now <a href="https://pulp.plan.io/issues/8805" target="_blank">https://pulp.plan.io/issues/8805</a> . In addition to resolving it with methods like ^, when 3.14 comes out (scheduled for June 29th) it would be great if you could report on if the improvements helped you.<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Tue, Jun 22, 2021 at 10:34 AM Bin Li (BLOOMBERG/ 120 PARK) <<a href="mailto:bli111@bloomberg.net" target="_blank">bli111@bloomberg.net</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div style="font-size:small;font-family:"Courier New",Courier,"BB.FixedWidth";white-space:pre-wrap">We recently add more clients to use the pulp content server. The processes run out the file descriptor first. We then increased both nginx and pulp-content by creating a override.conf <div>/etc/systemd/system/pulpcore-content.service.d # cat override.conf </div><div>[Service]</div><div>LimitNOFILE=65536</div><div><br><div>and updated nginx.conf</div><div># Gunicorn docs suggest this value.</div><div>worker_processes 1;</div><div>events {</div><div> worker_connections 10000; # increase if you have lots of clients</div><div> accept_mutex off; # set to 'on' if nginx worker_processes > 1</div><div>}</div><div><br></div><div>worker_rlimit_nofile 20000;</div><div><br></div><div><br></div><div>Now we are keep getting this error.</div></div><div>2021/06/22 11:26:36 [error] 78373#0: *112823 upstream timed out (110: Connection timed out) while connecting to upstream, client:</div><div><br></div><div>It looks like pulp-content server cannot keep up with requests. Is there anything we could do to increase the performance of the content server?</div></div>_______________________________________________<br>Pulp-list mailing list<br><a href="mailto:Pulp-list@redhat.com" target="_blank">Pulp-list@redhat.com</a><br><a href="https://listman.redhat.com/mailman/listinfo/pulp-list" target="_blank">https://listman.redhat.com/mailman/listinfo/pulp-list</a></blockquote></div> _______________________________________________<br>Pulp-list mailing list<br><a href="mailto:Pulp-list@redhat.com" target="_blank">Pulp-list@redhat.com</a><br><a href="https://listman.redhat.com/mailman/listinfo/pulp-list" target="_blank">https://listman.redhat.com/mailman/listinfo/pulp-list</a></blockquote></div></div></blockquote><br></div></div></div></div></div></div>_______________________________________________<br> Pulp-list mailing list<br> <a href="mailto:Pulp-list@redhat.com" target="_blank">Pulp-list@redhat.com</a><br> <a href="https://listman.redhat.com/mailman/listinfo/pulp-list" rel="noreferrer" target="_blank">https://listman.redhat.com/mailman/listinfo/pulp-list</a></blockquote></div> </blockquote></div>