python / psf-chef Goto Github PK
View Code? Open in Web Editor NEWChef configuration and cookbooks for the Python Software Foundation
Home Page: https://psf-chef.readthedocs.org/
Chef configuration and cookbooks for the Python Software Foundation
Home Page: https://psf-chef.readthedocs.org/
virt-kchn16.psf.osuosl.org and virt-vm43og no longer exists on the ganeti cluster, but are still noted in the opscode management panel.
should we clear these?
Should look something like this in the nginx config:
server {
listen 80;
server_name {{ server_ip }};
location / {
return 403;
}
}
Should be manage.py import_ics_calendars in the virtualenv.
Issue came through here: python/pythondotorg#76
I tend to agree. Thoughts? I feel like @dstuft mentioned doing this in VCL, but the last couple of days are a bit of a blur.
Hi,
We need the staging system to pull and use a different branch for production, otherwise it isn't really a "staging" system just a "won't be in production for a few minutes" system :)
Can someone help me get that setup?
https://www.ssllabs.com/ssltest/analyze.html?d=packages.python.org&latest
Subject | *.python.org
Fingerprint SHA256: 32a3b8d6138d2ad362e1b505e4df7eb57a4022a6e659df05c01324b56d8a628e
Pin SHA256: R3OdAsOX6cLLGc9VLDLaOo1U9AOfvgZ7iVH8gts6P9g=
-- | --
Common names | *.python.org
Alternative names | *.python.org python.org
Serial Number | 1e59fbd6a9f6a09bc72d22f2fb3c617b
Valid from | Thu, 02 Feb 2017 00:00:00 UTC
Valid until | Sun, 02 Feb 2020 23:59:59 UTC (expired 15 hours and 23 minutes ago) EXPIRED
Does the PSF have something like that setup? Right now I get flooded with error emails anytime the site goes down. Wondering if that exists for other projects and we just haven't gotten around to setting it up on python.org.
Hi!
Ok so in reference to this Python.org ticket (python/pythondotorg#505) PEP images still aren't displaying.
I believe it's because /srv/redesign.python.org/shared/media is owned root:root and cannot be written to by www-data which is what the cron and app code runs as. I'm honestly surprised an exception isn't raised when this happens, but apparently it isn't. Can we get this fixed asap?
Couple of things:
Just pushed a few things live, one of which had a migration. The migration was not run on staging or prod, had to do it manually.
Migration worked flawlessly on staging, but on Prod it complained about the peps app not having migrations. Had to comment out peps from INSTALLED_APPS to get the site working again. I'm guessing this is left over .pyc or perhaps pep related ghost migrations in the prod db?
Requirements for master (staging) and release (prod) both have Django==1.5.12 however requirements.txt on both boxes has 1.5.11 still. Cron delay issue maybe?
responds to pings, but I can't get a shell. hasn't had chef run in 3 months.
Found this earlier this morning: python/pythondotorg#56
There is no node for this in dyn, should there be? If so, where does it go?
Need MEDIA_ROOT setup for being able to upload images on preview and prod. Needs to be a directory setup outside of the repo (otherwise we'd lose media on each roll out).
And nginx needs to be setup to delivery the contents on the URL /m/.
MEDIA_ROOT will need to be set to that path in pydotorg.settings.staging.py
Should be writeable by the gunicorn process FYI.
We're changing how PEP Pages are generated to not break the existing PEP process around the Mercurial repo. The process is outlined here all that needs to happen is:
This assumes that PEP_REPO_PATH
has been set in the Django settings to the directory where the hq repository resides.
I would presume that this should be run hourly at most, perhaps just 4 times daily or even just daily.
Let me know if you have any questions!
Didn't dive into it too deeply, but I believe Fastly is caching /admin/ for a few seconds as I'm seeing very odd results that resolve themselves in under a minute. Can someone check on that and disable caching of /admin/*?
looks like this node has run out of inodes.
ernestd@virt-byl2f7:/tmp$ df -h
Filesystem Size Used Avail Use% Mounted on
/dev/vda3 8.9G 4.7G 3.8G 56% /
udev 489M 12K 489M 1% /dev
tmpfs 200M 260K 199M 1% /run
none 5.0M 0 5.0M 0% /run/lock
none 498M 0 498M 0% /run/shm
/dev/vda1 92M 55M 32M 64% /boot
ernestd@virt-byl2f7:/tmp$ df -i
Filesystem Inodes IUsed IFree IUse% Mounted on
/dev/vda3 588672 588672 0 100% /
udev 125047 448 124599 1% /dev
tmpfs 127376 331 127045 1% /run
none 127376 4 127372 1% /run/lock
none 127376 1 127375 1% /run/shm
/dev/vda1 24096 235 23861 1% /boot
It is setup for the evote role. @coderanger is this something we need to fix immediately, or is the evote thing on hold?
I'd say once daily at this point:
manage.py rebuild_index --noinput
FYI the --noinput is important as there is a Python 2 vs 3 bug that has been fixed, but not released yet.
This has been weighing heavily on my mind since I did it, so I need to let it out and confess. I didn't use chef's environments on purpose, because it was easier to just use an attribute.
It won't take a whole lot to get it using environments. I'll try to get to that this week at some point. This will also help solve things like #110
Tasks off the top of my head:
current_env
become node.chef_environment
see https://github.com/python/psf-chef/blob/master/cookbooks/pydotorg-redesign/recipes/default.rb#L4TBD: what versions on what env?
I tried to get a shell, but obviously couldn't get in since my user/key would never have made it to the server via Chef.
@coderanger could you poke around?
Hi Guys,
We're going to need/want migrations to be run when new code is released. It's "manage.py migrate" with the proper settings for the site.
Need to adjust Django's cookie domains on staging to be:
CSRF_COOKIE_DOMAIN = 'staging.python.org'
SESSION_COOKIE_DOMAIN = 'staging.python.org'
Otherwise we can't login to staging to poke at changes before releasing. Thanks!
We're going to need this to support the sites.
Should be manage.py update_blogs with the right virtualenv path.
Hi Everyone,
I think the Haystack and/or ElasticSearch configuration is borked on staging and prod. Ic an't get ANY results to show up on either system and can't rebuild the index on staging get connection errors.
Here is an example configuration for my local config:
HAYSTACK_CONNECTIONS = {
'default': {
'ENGINE': 'haystack.backends.elasticsearch_backend.ElasticsearchSearchEngine',
'URL': 'http://127.0.0.1:9200/',
'INDEX_NAME': 'haystack',
},
}
Can we get this updated asap? Thanks!
Hey,
So I'm guessing that prod got moved to a new box? Perhaps pydotorg-prod.iad1.psf.io or lb0.iad1.psf.io?
I no longer have access to this box (user frank) and can't fix the migration issue.
I was working on virt-q0dzqq.psf.osuosl.org (prod) and virt-l4es2w.psf.osuosl.org (stage) and it took me forever to figure out why the site worked fine directly, but not via Fastly.
So site is broken, fix that worked on old prod was:
Hi,
Going to be making a couple more dev sprints on python.org and need Jacob Burch to have ssh access to the systems since he'll be the one doing most of the work. You can find his keys here:
Got a request from @frankwiles. Opening an issue to track.
11:59 fwiles: [16:48:36] benjaminws: Is it possible to remove redirects that in place on preview but not prod? i.e. as we go to roll out some of the apps we ended up "hiding" yesterday?
11:59 ***: Playback Complete.
Hi Guys,
Need to remove the /dev/peps/ redirect to legacy.python.org/dev/peps to expose the now working dynamic /dev/peps. This actually rolled out to production a couple of weeks ago, but I forgot about the redirect.
Thanks!
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.