Create Dev and Stage Instance for Mozillians.org

RESOLVED FIXED

Status

mozilla.org Graveyard
Server Operations
RESOLVED FIXED
7 years ago
3 years ago

People

(Reporter: ozten, Assigned: jabba)

Tracking

Details

(Whiteboard: [allhands], URL)

(Reporter)

Description

7 years ago
After project kick-off, please create a stage instance for Mozillians.org.

== Date Needed By ==
* 2011/06/7

For Operations details, see:
https://mana.mozilla.org/wiki/display/websites/Mozillians.org

== HTTPS ==
* HTTPS is required for this site. (cert, IP address, etc)
(Reporter)

Updated

7 years ago
Blocks: 658621
Depends on: 665522
Depends on: 656189
Severity: enhancement → normal
Depends on: 665524

Updated

7 years ago
Assignee: server-ops → cshields
Depends on: 666932
Can we get an update on this?  Or at least an ETA?  This was filed 3 months ago, and Ozten asked that this be ready by June - clearly that wasn't possible, but there wasn't a new date provided.
(In reply to comment #1)
> Can we get an update on this?  Or at least an ETA?  This was filed 3 months
> ago, and Ozten asked that this be ready by June - clearly that wasn't
> possible, but there wasn't a new date provided.

The "Date Needed By" exceeded our available resources for getting it done.

The web side of this dev and stage is setup, but we are waiting on the ldap side which has a lot of dependencies right now.  (training, puppet modules, etc)

Depending on the training timeline with our contractor (which is coming soon I believe) I'll give an ETA of 8/12 for dev and stage.
Is there anything more we can do to make sure we either hit the 8/12 ETA or before?
(Assignee)

Updated

7 years ago
Depends on: 674450
(Assignee)

Comment 5

7 years ago
the -dev instance is kinda up, but there seems to be no css:

[Thu Aug 18 02:43:16 2011] [error] [client x.x.x.x] File does not exist: /data/www/mozillians-dev.allizom.org/mozillians/media/css/common-min.css, referer: https://mozillians-dev.allizom.org/en-US/
[Thu Aug 18 02:43:16 2011] [debug] mod_headers.c(768): headers: ap_headers_error_filter()
[Wed Aug 17 18:43:16 2011] [error] [client x.x.x.x] File does not exist: /data/www/mozillians-dev.allizom.org/mozillians/media/js/common-min.js, referer: https://mozillians-dev.allizom.org/en-US/

Also, the site isn't auto-updating quite yet, since we need to get the update script straightened out. Ping me on irc to work through these issues.
Are we still on track for the staging server to be up this week?
Can I please get access to bug 666932? I'd like to know what's blocking this.
(In reply to Aakash Desai [:aakashd] from comment #6)
> Are we still on track for the staging server to be up this week?

We are trying to get this up for you, but I don't recall ever promising it to be up this week?  Is the app ready for QA and Infrasec review?  If it is code complete then we are ready for stage.  Otherwise, there should be no rush.

In the meantime, while working on stage I am running in to migrations failing:

# python vendor/src/schematic/schematic migrations
Running migration 3:
DROP TABLE IF EXISTS invites; -- things we can do before we launch.

CREATE TABLE `invite` (
    `id` integer AUTO_INCREMENT NOT NULL PRIMARY KEY,
    `inviter` varchar(32) NOT NULL,
    `recipient` varchar(75) NOT NULL,
    `redeemer` varchar(32) NOT NULL,
    `code` varchar(32) NOT NULL UNIQUE,
    `redeemed` datetime,
    `created` datetime NOT NULL
) ENGINE=InnoDB CHARSET=utf8;

Error: Had trouble running this: BEGIN;
DROP TABLE IF EXISTS invites; -- things we can do before we launch.

CREATE TABLE `invite` (
    `id` integer AUTO_INCREMENT NOT NULL PRIMARY KEY,
    `inviter` varchar(32) NOT NULL,
    `recipient` varchar(75) NOT NULL,
    `redeemer` varchar(32) NOT NULL,
    `code` varchar(32) NOT NULL UNIQUE,
    `redeemed` datetime,
    `created` datetime NOT NULL
) ENGINE=InnoDB CHARSET=utf8;

UPDATE schema_version SET version = 3;
COMMIT;
stdout: 
stderr: ERROR 1050 (42S01) at line 4: Table 'invite' already exists

returncode: 1
(In reply to Corey Shields [:cshields] from comment #8)
> (In reply to Aakash Desai [:aakashd] from comment #6)
> > Are we still on track for the staging server to be up this week?
> 
> We are trying to get this up for you, but I don't recall ever promising it
> to be up this week?  Is the app ready for QA and Infrasec review?  If it is
> code complete then we are ready for stage.  Otherwise, there should be no
> rush.

The app is ready for QA and Infrasec review and the former has already begun their testing on the dev-server. We'd like to open testing up to a wider audience and a stage server would help that out immensely...and quickly because the window for releasing this app this quarter is at about 2-3 weeks.

It sounds like there's some miscommunication here though, so I'll end with a question: What's stopping the staging server from being up by today (other than the migrations)?
(In reply to Aakash Desai [:aakashd] from comment #9)
> It sounds like there's some miscommunication here though, so I'll end with a
> question: What's stopping the staging server from being up by today (other
> than the migrations)?

Time..  Getting the stage ldap nodes done.  Getting their network flows opened up.  260 other server ops bugs.  :)

and the migrations issue

thanks!
(In reply to Corey Shields [:cshields] from comment #10)
> (In reply to Aakash Desai [:aakashd] from comment #9)
> > It sounds like there's some miscommunication here though, so I'll end with a
> > question: What's stopping the staging server from being up by today (other
> > than the migrations)?
> 
> Time..  Getting the stage ldap nodes done.  Getting their network flows
> opened up.  260 other server ops bugs.  :)
> 
> and the migrations issue
> 
> thanks!

Can we do better on timing and set on a date on when this can be up and going?

Folks are working on the dev-server right now and that satisfies all our needs, but I really want to get rid of all of our IT blockers (i.e. staging and production nodes) by the end of this week or mid-next week. Help a brutha out?
(Assignee)

Comment 12

7 years ago
I've got netops working on the flows right now. I still need the site to go through the initial code review before we can really set much else up. LDAP backed sites are new territory and we need to invent some things to make it work in a scalable fashion. Currently the code has the webapp binding to ldap as the root user. This needs to be changed. There also needs to be an admin user whose credentials are known by the admin host but not by the webheads. We don't have anything in place for this yet and we still need to come up with something. There are still some technical things to deal with. I am working on it full time today and this has the highest priority for me right now.
(In reply to Aakash Desai [:aakashd] from comment #11)
> Can we do better on timing and set on a date on when this can be up and
> going?
> 
> Folks are working on the dev-server right now and that satisfies all our
> needs, but I really want to get rid of all of our IT blockers (i.e. staging
> and production nodes) by the end of this week or mid-next week. Help a
> brutha out?

Sure, I'll guarantee it to be done by 9/23 but in reality I hope to have it done in a matter of hours/days.  I am speaking to this specific bug, dev and stage.  Prod is another issue, we have not yet discussed how to handle ldap on the prod side.

I'm not trying to be crass, you are in line with hundreds of other "Help a brutha out" requests.  Literally.  Some of them are pretty urgent and have to do with services that are up and running today and need help, let alone the new services we need to turn up.

I can say "I hope to have it done in hours" and then break that promise when a service dies and we have to react to it, or a blocking web push comes in from another web service, etc..  And I hate to do that.  We are working on this as we speak.  Help a brutha out and fix the migrations issue so we are not blocked on that.  :)

My team is currently down 2 people.  Getting to these as quickly as possible.  Thanks for understanding.
> We are working on this as we speak.  Help a brutha out and fix the migrations issue > so we are not blocked on that.  :)

Dave, Ozten: can you help Corey out with the migration issues?
(Reporter)

Comment 15

7 years ago
Sorry, we missed this bug until now. Looking into it...
Hey Corey, 

I've been out of town and allhands is eating my time, so feel free to ping me if we're being unresponsive.

I did this which should unblock you:

http://github.com/mozilla/mozillians/commit/1cf4df2


but also we may want to add rmpyc as part of our build process.  I didn't think we were making pyc files in our environments, but oremj knows more about that than I.  It definitely trips us up enough that we should rm all pyc files during an update.
web side of mozillians stage is done, waiting for ldap side.
Assignee: cshields → jdow
Whiteboard: [allhands]
(Reporter)

Comment 18

7 years ago
We've made it easier to deploy ldap files to one host, instead of the webcode model in Bug#686697.
Depends on: 686697
(Assignee)

Comment 19

7 years ago
dev and stage are online. Prod shouldn't be too far away, once infrasec and qa has signed off. We'll need to get some hosts built for the LDAP backend for prod, but it is all working with puppet now, so deployment will be quick and easy.
Status: NEW → RESOLVED
Last Resolved: 7 years ago
Resolution: --- → FIXED
Where is this staging server?
(at what URL)
(In reply to Aakash Desai [:aakashd] from comment #20)
> Where is this staging server?

Dev: https://affiliates-dev.allizom.org
Stage: https://affiliates.allizom.org

Unless I'm mistaken.
Stephend with the inspiration!

https://mozillians.allizom.org , Thanks.
(In reply to Stephen Donner [:stephend] from comment #22)

> Unless I'm mistaken.

Which I am, so:

Dev: https://mozillians-dev.allizom.org
Stage: https://mozillians.allizom.org
Product: mozilla.org → mozilla.org Graveyard
You need to log in before you can comment on or make changes to this bug.