Migrate netops-apps.mozilla.org to phx1 generic cluster



Infrastructure & Operations
WebOps: Other
5 years ago
4 years ago


(Reporter: jd, Assigned: phrawzty)



(Whiteboard: [triaged 20121121][push interrupt], URL)



5 years ago
Need to migrate netops-apps.mozilla.org from intranet1.webapp.scl3.mozilla.com (old mrapp-intranet01) to phx1 generic cluster.

This site is broken as it is in development, as such we need all 3 environs but to leave prod empty until there is a working site to put in place.

Ravi is the contact for any questions.
Blocks: 772996

Comment 1

5 years ago
netops-apps.m.o needs to write to a mysql database.

The the quagga backend (running on netops2) needs to be able to query the MySql server every few minutes.

As I understand it this means there needs to be a flow from netoos(1|2).private.scl3.mozilla.com 10.22.75.(126|127) to generic-rw-zeus.db.phx1.mozilla.com (

Just adding notes here and will open a separate bug for the flows.

Also the new database name will be like netops_apps_mozilla_org-banhammer as there will be several apps running under the netops-apps domain eventually.


5 years ago
Depends on: 811455

Comment 2

5 years ago
This is now complete, there are also dev and stage instances now:


These are all behind LDAP auth like the original. I have set the dev site to auto-update every 10 min.

I created a mana page for it, please update it with your development docs and whatever else might be useful. https://mana.mozilla.org/wiki/display/websites/netops-apps.mozilla.org

There are two issues you might like to address at some point.

First the settings.py file is checked into the SVN repo. I have edited that file locally for all 3 instances (dev / stage / prod) and it is fine for now but it would be far better to have that file include a local-settings.py file and the I can migrate my settings there.

The second issue is that the site does not actually work (I am told you are aware of this already) and I will include the log output from one such error below.

Let me know if you need anything more from me.


[Fri Nov 16 14:47:18 2012] [info] mod_wsgi (pid=12783): Create interpreter 'netops-apps-dev.allizom.org|'.
[Fri Nov 16 14:47:18 2012] [info] [client] mod_wsgi (pid=12783, process='netops-apps-dev-ssl', application='netops-apps-dev.allizom.org|'): Loading WSGI script '/data/www/netops-apps-dev.allizom.org/netops-apps/apache/django.wsgi'.
[Fri Nov 16 14:47:18 2012] [error] [client] mod_wsgi (pid=12783): Target WSGI script '/data/www/netops-apps-dev.allizom.org/netops-apps/apache/django.wsgi' cannot be loaded as Python module.
[Fri Nov 16 14:47:18 2012] [error] [client] mod_wsgi (pid=12783): Exception occurred processing WSGI script '/data/www/netops-apps-dev.allizom.org/netops-apps/apache/django.wsgi'.
[Fri Nov 16 14:47:18 2012] [error] [client] Traceback (most recent call last):
[Fri Nov 16 14:47:18 2012] [error] [client]   File "/data/www/netops-apps-dev.allizom.org/netops-apps/apache/django.wsgi", line 9, in <module>
[Fri Nov 16 14:47:18 2012] [error] [client]     import django.core.handlers.wsgi
[Fri Nov 16 14:47:18 2012] [error] [client] ImportError: No module named django.core.handlers.wsgi
Last Resolved: 5 years ago
Resolution: --- → FIXED
So the issue here is that django isn't installed system wide on the generic webheads. The architecture expects a plahdoh app with a vendor directory housing django and any other site specific dependencies. So we there are 2 options.

1. Move the site to a place with django installed system wide.
2. Port the application to playdoh.
Here is option 2 : https://github.com/XioNoX/BanHammer can you update it?
Assignee: jcrowe → server-ops-webops
Resolution: FIXED → ---

Comment 5

5 years ago
06:54:46 < solarce> phrawzty: /data/genericrhel6-dev/src/netops-apps-dev.allizom.org/netops-apps-xionix/
06:55:04 < solarce> phrawzty: i cloned the repo, got the submodule setup and made local.py
06:55:15 < solarce> phrawzty: local.py should be reviewed for any other bits that need turned on

:solarce has handed me the keys, and i'll drive the rest of the way. :)
Assignee: server-ops-webops → dmaher
Priority: -- → P4
Whiteboard: [triaged 20121121][push interrupt]
Also moved the new directory into place and did a deploy, Dan the Man will fix up Apache

Comment 7

5 years ago
Apache is all fixed up; just need to work on getting the required dependencies into place properly.

[Wed Nov 21 07:09:15 2012] [error] [client] ImportError: No module named netaddr
Dependency added the the latest commit (in the local-vendor folder).

Comment 9

5 years ago
Everything appears to work _except_ for the static assets :

09:07:26 < XioNoX> phrawzty: okay, I know how to fix that
09:08:13 < XioNoX> it's ok, I'll do that later

Waiting for :xionox to fix before closing bug.
It's now fully working in dev. Move to prod is tracked in bug 814135.
Last Resolved: 5 years ago5 years ago
Resolution: --- → FIXED
Component: Server Operations: Web Operations → WebOps: Other
Product: mozilla.org → Infrastructure & Operations
You need to log in before you can comment on or make changes to this bug.