Status

Infrastructure & Operations
WebOps: IT-Managed Tools
RESOLVED WONTFIX
5 years ago
3 years ago

People

(Reporter: lonnen, Unassigned)

Tracking

Details

(Whiteboard: [kanban:https://webops.kanbanize.com/ctrl_board/2/48] [triaged 20121214][waiting][netops])

(Reporter)

Description

5 years ago
Not sure if I've filed this in the right component, please move around as necessary.

Where can I set up a flask app with no data store that will be able to send and receive requests to both Github and WebQA's Jenkins server (behind MV VPN)?


WebDev has been using a small flask app + Jenkins to test PRs automatically pre-merge. WebQA has asked me to set up the same bot for their B2G tests, which means it needs access to their Jenkins in order to run the tests on their mobile hardware.

Comment 1

5 years ago
Normally I would think something like this fits on our generic cluster. The github side should be fine. However needing access to MV-VPN concerns me... I don't know if/how we can make that work.

Let's get some NetOps advice here.

@NetOps: web app, want to put it on generic cluster in PHX1 (webapp VLAN), needs access to github and a host in MV (http://qa-selenium.mv.mozilla.com:8080/). Can we do this?


@lonnen: We have a bug to eventually merge WebQA's Jenkins with ci.mozilla.org... bug 693544. No idea when that might happen, it's a big project and not very high on the priority list right now.
Flags: needinfo?(ravi)
Priority: -- → P3
Whiteboard: [triaged 20121214][waiting][netops]
(Reporter)

Comment 2

5 years ago
@jakem: thanks for looking at this.

I don't know where it fits into the rest of the priorities. I imagine merging a public CI with webQAs restricted CI won't be sorted for a while, yet. Thanks for looking into this.

Comment 3

5 years ago
The two environments are separate and we try to keep it that way.  Generally we would ask that you move whatever resources that are in MV to the data center since an office SLA is different from a data center (no redundant power, different hardware, etc) so we can't guarantee uptime.  We also keep the environments separate because access control policies are much easier managed and predictable in the data center than they are in the office.  Once you start talking about the office hosts connecting to the data center (or vice versa) it becomes a slippery slope fairly fast.

So we can technically accommodate the flow, but no matter what we may be entering Opsec territory and I've Cc'd them.
Flags: needinfo?(ravi) → needinfo?(mpurzynski)

Updated

5 years ago
Priority: P3 → --
I have to agree with Ravi here. There are reasons we keep things this way. Let's move the host into a data center and keep things the way they should be - separated.
Flags: needinfo?(mpurzynski)
Component: Server Operations: Web Operations → WebOps: IT-Managed Tools
Product: mozilla.org → Infrastructure & Operations
Depends on: 693544
Whiteboard: [triaged 20121214][waiting][netops] → [kanban:https://kanbanize.com/ctrl_board/4/90] [triaged 20121214][waiting][netops]

Updated

4 years ago
Whiteboard: [kanban:https://kanbanize.com/ctrl_board/4/90] [triaged 20121214][waiting][netops] → [kanban:engops:https://mozilla.kanbanize.com/ctrl_board/6/2510] [kanban:https://kanbanize.com/ctrl_board/4/90] [triaged 20121214][waiting][netops]
Whiteboard: [kanban:engops:https://mozilla.kanbanize.com/ctrl_board/6/2510] [kanban:https://kanbanize.com/ctrl_board/4/90] [triaged 20121214][waiting][netops] → [kanban:https://kanbanize.com/ctrl_board/4/90] [triaged 20121214][waiting][netops]

Updated

3 years ago
Whiteboard: [kanban:https://kanbanize.com/ctrl_board/4/90] [triaged 20121214][waiting][netops] → [kanban:https://webops.kanbanize.com/ctrl_board/2/48] [triaged 20121214][waiting][netops]
(Reporter)

Comment 5

3 years ago
Is this still desired?
Status: NEW → RESOLVED
Last Resolved: 3 years ago
Resolution: --- → WONTFIX
(Reporter)

Comment 6

3 years ago
ugh. accidental resolution switch.

@stephend -- do you still want this?
Status: RESOLVED → REOPENED
Flags: needinfo?(stephen.donner)
Resolution: WONTFIX → ---
(In reply to Chris Lonnen :lonnen from comment #6)
> ugh. accidental resolution switch.
> 
> @stephend -- do you still want this?

I'll let Dave chime in here, but: Mozilla.org and Web QA would <3 it if we could get something like this working for either/both Travis-CI and Jenkins
Flags: needinfo?(stephen.donner) → needinfo?(dave.hunt)
I'm going to mark this as WONTFIX because we'll probably solve this in a different way. For Travis-CI there's no issue (it can already run GitHub pull requests). For Jenkins, we're currently in the process of setting up a publicly accessible instance dedicated for Web QA. As this will be public we should be able to set up a limited LDAP user that GitHub can use in a webhook for triggering builds.
Status: REOPENED → RESOLVED
Last Resolved: 3 years ago3 years ago
Flags: needinfo?(dave.hunt)
Resolution: --- → WONTFIX
You need to log in before you can comment on or make changes to this bug.