Awesome! One thing to note is that if download_concurrency was a problem, the problem may temporarily go away if there’s nothing (or very little) new packages to download. If a new major update is released and 100s of packages are updated, you might hit it again. Keep that in mind.
@Justin_Sherrill
So is the lower the number for download_concurrency better? I would think that the higher the number would be mean better performance.
It does mean better performance but it also means the higher likelyhood that the server will start rejecting your requests. Pulp 2 was a lot more aggressive with retrying downloads. Pulp3 is a lot more conservative in this regard and will fail a sync at the first hint of trouble 
Could you help me with the process on implementing this change on a CentOS 7 installation? I’m not sure exactly where I need to add this option for the ‘Bad Gateway’ error messages.
Thank you.
Which change exactly? the download concurrency change?
Also, i started seeing the same issue with syncs and filed an issue here: Issue #8867: downloads from cdn.redhat fail with 'Cannot connect to host cdn.redhat.com:443 ssl:default [[SSL: SSLV3_ALERT_UNEXPECTED_MESSAGE] sslv3 alert unexpected message (_ssl.c:877)]' - Pulp
@Justin_Sherrill I was looking at the change for implementing the disableuse=no for the pulp core-content service. However I’m having some big issues after applying the updates today. When running any commands to get the status of the services from ‘foreman-maintain service status’ I receive this message.
Running Status Services
Get status of applicable services:
Displaying the following service(s): [FAIL]
undefined method `enabled?’ for RemoteDB(rh-mongodb34-mongod [5]):ForemanMaintain::Utils::Service::RemoteDB
Scenario [Status Services] failed.
The following steps ended up in failing state:
[service-status]
Resolve the failed steps and rerun
the command. In case the failures are false positives,
use --whitelist=“service-status”
Both MongoDB and PostgreSQL are running on a remote server.
hrmm, what version of rubygem-foreman_maintain is installed? sounds like its not taking the ‘remote’ nature into consideration
After updating today it is running rubygem-foreman_maintain-0.7.9-1.el7.noarch. The previous version was rubygem-foreman_maintain-0.7.5-1.el7.noarch. I am able to reproduce in both non-prod and prod environments.
Hi folks!
I faced the same issues on different kind of environment: azure, local kvm… etc just for the sake of testing. I was completely unable to sync any “large” repository… or the sync will just crash at 80% or so…
In the end, it seems this issue disappeared when I gave my VMs a little bit more cpu and memory.
My initial assumption was that those requirements were probably for production grade deployment… I was wrong (and it’s stated in the documentation)… anyway, before going too deep into the code, make sure you meet those 
Any resolution on this? I recently apply updates in our test/qa environment to bring it to 2.3.5 and also migrated the pulpcore database to the environments dedicated postgresql server. Now whenever I try to work with foreman services, I get the error ‘undefined method `enabled?’ for RemoteDB(postgresql:candlepin [10]):ForemanMaintain::Utils::Service::RemoteDB’ for any service related command in foreman-maintain unless I use the switch ‘–exclude postgresql’. I would like to have a solution to this issue before I apply updates and do the same DB migration of pulpcore in our production environment.
Sorry for the late reply, but I haven’t heard of any resolution yet. I still have the same issue with my Katello environment split with the application and data bases on different hosts. I have not upgraded to version 4 yet as I’m still running the latest release of 3.18.
here’s the bug: Bug #32319: service stop with remote databases fails - Foreman Maintain - Foreman
it is resolved but may not be backported to your given release. I would try a
yum update http://yum.theforeman.org/nightly/el7/x86_64/rubygem-foreman_maintain-0.8.9-1.el7.noarch.rpm
that will update to the newest foreman_maintain that has this fix.
1 Like
Also for the download issues, this is resolved in pulp 3.14 by adding retry support . Upgrading to katello 4.1.2 should get that for you.
2 Likes