Bugzilla will be upgraded to version 5.0. The upgrade date is tentatively scheduled for 2 December 2018, pending final testing and feedback.
Bug 1449743 - Add session-max-unacked option to installer
Add session-max-unacked option to installer
Status: CLOSED ERRATA
Product: Red Hat Satellite 6
Classification: Red Hat
Component: Installer (Show other bugs)
Nightly
x86_64 Linux
high Severity urgent (vote)
: 6.2.11
: Unused
Assigned To: Chris Roberts
Roman Plevka
http://projects.theforeman.org/issues...
: Performance, PrioBumpField, Triaged, UserExperience
Depends On:
Blocks:
  Show dependency treegraph
 
Reported: 2017-05-10 10:59 EDT by Chris Roberts
Modified: 2018-08-31 11 EDT (History)
6 users (show)

See Also:
Fixed In Version: katello-installer-base-3.0.0.89-1
Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of:
: 1463807 (view as bug list)
Environment:
Last Closed: 2017-08-10 13:02:29 EDT
Type: Bug
Regression: ---
Mount Type: ---
Documentation: ---
CRM:
Verified Versions:
Category: ---
oVirt Team: ---
RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: ---


Attachments (Terms of Use)


External Trackers
Tracker ID Priority Status Summary Last Updated
Apache JIRA QPID-7677 None None None 2017-05-10 10:59 EDT
Foreman Issue Tracker 19514 None None None 2017-05-10 12:01 EDT
Red Hat Product Errata RHBA-2017:2466 normal SHIPPED_LIVE Satellite 6.2.11 Async Release 2017-08-10 17:01:20 EDT

  None (edit)
Description Chris Roberts 2017-05-10 10:59:13 EDT
FYI the upstream patch for https://issues.apache.org/jira/browse/QPID-7677 adds new configurable option to manage size of outgoing buffer of unsettled/unacknowledged messages. The bigger the buffer is, the more messages can be pending to the consumer but the bigger memory requirements are there.

We should decide what will be the value for Satellite.

Satellite consumers are:
- katello-agent/goferd 
  - very few messages pending at once (usually just 1)
- pulp celery workers/manager (on 0.10 now, though unaffected until upgraded to 1.0)
  - I *think* limiting capacity to say 10 or 20 wont be problem even for resource_manager
- celery monitoring its event queue
  - here the bandwidth might matter the most / this is the most frequent AMQP subscription - so far on 0.10 so unaffected
- pulp.task queue (where http wsgi script writes requests to pulp)
  - also quite busy, less than "celery" queue, shall be ok
- various auxiliary queues, usually no deal wrt throughput

So my suggestion would be to add

session-max-unacked=10

to /etc/qpid/qpidd.conf in final solution/deployment/installer.
Comment 1 Chris Roberts 2017-05-10 11:02:20 EDT
http://projects.theforeman.org/issues/19514
Comment 3 pm-sat@redhat.com 2017-05-10 14:02:12 EDT
Upstream bug assigned to chrobert@redhat.com
Comment 4 pm-sat@redhat.com 2017-05-10 14:02:14 EDT
Upstream bug assigned to chrobert@redhat.com
Comment 7 Roman Plevka 2017-08-02 07:45:52 EDT
Did this got in for snap2?
I can't find such installer option.

Also, i can't even see such option in qpid.conf

# satellite-installer -h | grep unacked
# rpm -qa satellite
satellite-6.2.11-2.0.el7sat.noarch
# rpm -qa katello
katello-3.0.0-21.el7sat.noarch
# rpm -qa qpid*
qpid-tools-0.34-2.el7.noarch
qpid-cpp-server-0.34-25.el7sat.x86_64
qpid-proton-c-0.9-16.el7.x86_64
qpid-qmf-0.34-4.el7.x86_64
qpid-cpp-client-devel-0.34-25.el7sat.x86_64
qpid-dispatch-router-0.4-22.el7sat.x86_64
qpid-cpp-client-0.34-25.el7sat.x86_64
qpid-cpp-server-linearstore-0.34-25.el7sat.x86_64
Comment 8 Chris Roberts 2017-08-03 12:30:30 EDT
Opened a new PR against this to expose the options to the installer for the user to configure and us ship the options set in the BZ... moving back to POST since a new PR
Comment 11 jcallaha 2017-08-09 09:32:25 EDT
Verified in Satellite 6.2.11 Snap 4. The installer now correctly shows the two new options and they are stored in the appropriate configuration file.


[root@smqa-r210-02 ~]# satellite-installer --help | grep qpid-
...
    --katello-qpid-session-unacked  Buffer if the broker has a large number of sessions and the memory overhead is a problem (default: "-10")
    --katello-qpid-wcache-page-size  The size (in KB) of the pages in the write page cache (default: "4")


--- defaults ---

[root@smqa-r210-02 ~]# tail -10 /etc/qpid/qpidd.conf
ssl-port=5671
ssl-cert-db=/etc/pki/katello/nssdb
ssl-cert-password-file=/etc/pki/katello/nssdb/nss_db_password-file
ssl-cert-name=broker


session-max-unacked=-10

wcache-page-size=4


--- changing values via installer ---

[root@smqa-r210-02 ~]# satellite-installer --katello-qpid-session-unacked 100 --katello-qpid-wcache-page-size 25
Installing             Done                                               [100%] [..........................................................................................................]
  Success!
  * Satellite is running at https://smqa-r210-02.lab.eng.brq.redhat.com
  * To install additional capsule on separate machine continue by running:

      capsule-certs-generate --capsule-fqdn "$CAPSULE" --certs-tar "~/$CAPSULE-certs.tar"

  The full log is at /var/log/foreman-installer/satellite.log


[root@smqa-r210-02 ~]# tail -10 /etc/qpid/qpidd.conf
ssl-port=5671
ssl-cert-db=/etc/pki/katello/nssdb
ssl-cert-password-file=/etc/pki/katello/nssdb/nss_db_password-file
ssl-cert-name=broker


session-max-unacked=100

wcache-page-size=25


--- qpid is then up and running fine ---

[root@smqa-r210-02 ~]# systemctl status qpidd
● qpidd.service - An AMQP message broker daemon.
   Loaded: loaded (/usr/lib/systemd/system/qpidd.service; enabled; vendor preset: disabled)
   Active: active (running) since Wed 2017-08-09 15:28:05 CEST; 1min 32s ago
     Docs: man:qpidd(1)
           http://qpid.apache.org/
 Main PID: 25411 (qpidd)
   CGroup: /system.slice/qpidd.service
           └─25411 /usr/sbin/qpidd --config /etc/qpid/qpidd.conf

Aug 09 15:28:05 smqa-r210-02.lab.eng.brq.redhat.com systemd[1]: Started An AMQP message broker daemon..
Aug 09 15:28:05 smqa-r210-02.lab.eng.brq.redhat.com systemd[1]: Starting An AMQP message broker daemon....
Comment 13 errata-xmlrpc 2017-08-10 13:02:29 EDT
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHBA-2017:2466

Note You need to log in before you can comment on or make changes to this bug.