Closed Bug 853185 Opened 11 years ago Closed 11 years ago

mdn: add ES settings for dev, stage, and prod settings_local.py

Categories

(Infrastructure & Operations Graveyard :: WebOps: Other, task, P4)

x86
macOS

Tracking

(Not tracked)

RESOLVED FIXED

People

(Reporter: groovecoder, Assigned: dmaher)

References

Details

(Whiteboard: [triaged 20130322])

Now that bug 824065 is done and we have pushed ES code, we can do this; slowly, server-by-server, and to verify each server. The settings are:

ES_DISABLED = False
ES_URLS = ['http://{dev,stage,prod es cluster host}:{dev, stage, prod es cluster port}']
ES_INDEXES = {'default': 'main_index'}
ES_INDEX_PREFIX = 'mdn'
ES_LIVE_INDEX = True
ES_INDEXING_TIMEOUT = 30

Let's do the dev server first.
Assignee: server-ops-webops → dmaher
Priority: -- → P4
Whiteboard: [triaged 20130322]
I have added the following lines to settings_local.py on Dev (only) :

# bug 853185
ES_DISABLED = False
ES_INDEXES = {'default': 'main_index'}
ES_INDEX_PREFIX = 'mdn'
ES_LIVE_INDEX = True
ES_INDEXING_TIMEOUT = 30
ES_URLS = ['http://elasticsearch-zlb.dev.vlan81.phx1.mozilla.com:9200']
Status: NEW → ASSIGNED
Something seems not right? https://developer-dev.allizom.org/en-US/ hangs forever now.

curl from the box works though ...

$ curl http://elasticsearch-zlb.dev.vlan81.phx1.mozilla.com:9200
{
  "ok" : true,
  "status" : 200,
  "name" : "elasticsearch2_dev_phx1",
  "version" : {
    "number" : "0.20.5",
    "snapshot_build" : false
  },
  "tagline" : "You Know, for Search"
}

James or Will - do these settings look right?
What is https://developer-dev.allizom.org/en-US/ doing? Is it doing a search or something else that's ES-related?
No, even the home page (no ES activity) is borked. :(
Blocks: 854459
The first error message in apache logs is/was:

[Sun Mar 24 09:48:21 2013] [error] [client 10.22.81.213] Script timed out before returning headers: kuma.wsgi

That's well after we deployed the ES_* changes, so maybe it's un-related? Did anything happen on 2013-03-24? Maybe just needs a reboot?
Hello,

Just a quick ping as there hasn't been any movement on this bug in a while.  Is this bug still valid ?  Is there anything I can help with ?
Flags: needinfo?
Dev looks good. We're going to iterate on our ES index design before we start using in stage or prod.
Flags: needinfo?
Whiteboard: [triaged 20130322] → [triaged 20130322][waiting dev]
Blocks: 868506
We've worked thru some ES index issues on dev. Can we make these changes to the stage server now? Do we have an ES stage cluster?
(In reply to Luke Crouch [:groovecoder] from comment #8)
> Do we have an ES stage cluster?

There is no generic ES stage cluster - functionally speaking, other projects simply use the dev ES cluster for staging as well (using a different index).
Okay. Let's make sure to set ES_INDEX_PREFIX = 'mdnstage' in the settings_local.py on stage so it doesn't re-use dev's index. Maybe we should even make ES_INDEX_PREFIX = 'mdndev' for dev too?
Luke: That's what we do with both SUMO and Input.
I've set this for dev :
ES_INDEX_PREFIX = 'mdndev'

And this for stage :
ES_INDEX_PREFIX = 'mdnstage'
We need more (all) of the ES_* settings to make stage work.
All of the ES_* settings exist in stage (see bug 824065#c19).
Stage looks good. Let's do prod now.
The following config block has been added on the admin host :

ES_DISABLED = False
ES_INDEXES = {'default': 'main_index'}
ES_INDEX_PREFIX = 'mdnprod'
ES_LIVE_INDEX = True
ES_INDEXING_TIMEOUT = 30
ES_URLS = ['http://elasticsearch-zlb.webapp.scl3.mozilla.com:9200']

You will need to trigger a push to prod for this to become active.
Whiteboard: [triaged 20130322][waiting dev] → [triaged 20130322]
Prod push, reindex, and search looks good. Thanks!
Status: ASSIGNED → RESOLVED
Closed: 11 years ago
Resolution: --- → FIXED
Component: Server Operations: Web Operations → WebOps: Other
Product: mozilla.org → Infrastructure & Operations
Product: Infrastructure & Operations → Infrastructure & Operations Graveyard
You need to log in before you can comment on or make changes to this bug.