On 10 June 2015 at 10:53, Dan Andreescu <dandreescu(a)wikimedia.org> wrote:
I see three ways for data to get into the cluster:
1. request stream, handled already, we're working on ways to pump the data
back out through APIs
Awesome, and it'd end up in the Hadoop cluster in a table? How...do we
kick that off most easily?
2. Event Logging. We're making this scale arbitrarily by moving it to
Kafka. Once that's done, we should be able to instrument pretty much
anything with Event Logging
3. Piwik. There is a small but growing effort to stand up our own piwik
instance so we can get basic canned reports out of the box and not have to
reinvent the wheel for every single feature we're trying to instrument and
learn about. This could replace a lot of the use cases for Event Logging
and free up Event Logging to do more free-form research rather than cookie
cutter web analytics.
So I'm asking, I guess, two things. The first
is: can we have a firm
commitment that we'll get this kind of stuff into Hadoop? Right now we
have a RESTful API everywhere that is not (to my knowledge) throwing
data into the request logs. We have a WDQS that isn't either.
Undoubtedly we have other tools I haven't encountered. It's paramount
that the first question we ask with new services or systems is "so
when does new traffic data start hitting the analytics cluster?"
The commitment has to be made on both sides. The teams building the
services have to instrument them, picking either 2 or 3 above. And then
we'll commit to supporting the path they choose. The piwik path may be slow
right now, fair warning.
Second: what's best practices for this? What
resources are available?
If I'm starting a service on Labs that provides data to third-parties,
What exactly do you mean here? That's a loaded term and possibly against
An API, Dan ;)
what would analytics recommend my easiest path is to getting request
logs into Hadoop?
Weighing everything on balance, right now I'd say adding your name to the
piwik supporters. So far, off the top of my head, that list is:
* wikimedia store
* annual report
* the entire reading vertical
* russian wikimedia chapter (most likely all other chapters would chime in
* a bunch of labs projects (including wikimetrics, vital signs, various
How is piwik linked to Hadoop? I'm not asking "how do we visualise the
data" I'm asking how we get it into the cluster in the first place.
Analytics mailing list