2,885 karma · joined August 28, 2009
formerly: ScaleFT CTO (acquired by Okta), cloudkick dude, rackspace racker. (yc w09, acquired dec 2010 by rackspace)
apache httpd hacker.
blog thing: https://paul.querna.org/
We had to write many things. There is a list of our dependencies at the bottom of this post:
http://journal.paul.querna.org/articles/2011/12/17/technolog...
Having said that, the only ones I consider moderately painful is when we need to deal with node-thrift, which on the Python or JVM is well maintained by existing groups.
Big Brother is pretty cool. I remember using it long ago too :)
I certainly respect the simplicity of BigBrother, but the scopes of the products are completely different. Rackspace Cloud Monitoring is a SaaS built with the ability to be used by hundreds of thousands of unique customers, monitoring millions of 'things'. It also has a goal of never losing the trust of our users -- user's generally don't care if your data center is having a bad day, but they certainly do care if you didn't alert them when their site is down, and so multi-region durability gets added to a requirement list.
Complexity does come from these kind of requirements.
The C client isn't what I would call great, and the API surface is fairly complicated for languages to bind to -- you see many reports about the bindings crashing their respective languages.
I'm currently hacking on a pure-Node.js implementation of the Apache Jute RecrodIO and Apache Zookeeper network protocol: https://github.com/racker/node-zk
The core problem is that Zookeeper is using a deprecated serialization layer as its network protocol. Jute predates Apache Thrift or Apache Avro, and it only has code-gen for C and Java. Ideally Zookeeper would add a second network protocol, one which was either a spec for a nice binary protocol, or a more modern serialization framework like Thrift or Avro, which support many more languages.
The ticket to change out Jute is here: https://issues.apache.org/jira/browse/ZOOKEEPER-102
But no one seems to actively be working on it.
Right now, Adobe has not even submitted a proposal to the incubator -- just blog posts about how they intend to submit it.
I eagerly await their submission of a Flex proposal. Part of the proposal, and critical for Flex, is proving this isn't just a code dump. If it is a code dump, I will vote -1, and I hope everyone else does too.
(can't edit anymore to make em URL shortened)
Rackspace is looking for various roles right now in our new location in SOMA (previously we were in the pre-acquisition Cloudkick Office):
* Node.js: For some really sweet distributed systems, looking for good systems programmers who either already hack on node, or can pick it up quickly. (No, no one has 5 years of node.js experience, but using Twisted, EventMachine, Java NIO, or lots of C are good introductions to things you need)
http://jobs.rackspace.com/job/San-Francisco-Sr-Python-Develo...
* Frontend Javascript: We are developing Rackspace's next generation of UIs for our customers. We need great frontend engineers, with a heavy focus on javascript to help us.
http://jobs.rackspace.com/job/San-Francisco-Front-End-Web-De...
* Python Hacking: We use a mixture of Django and Twisted Python, utilizing many interesting technologies like Libcloud and RabbitMQ.
http://jobs.rackspace.com/job/San-Francisco-Python-Developer...
More general info on our location: http://rackertalent.com/san-francisco/
If interested, check out the links above, or drop me a line, paul.querna@rackspace.com
In this rant he didn't say Node.js was the solution, or better than any of these crappy abstractions.
Don't assume that. He might not say he doesn't like node.js, but it doesn't mean he is happy with it.
re: 2) There is a product called RackConnect, which can bridge the public cloud which has API based provisioning, and physical servers: http://www.rackspace.com/hosting_solutions/hybrid_hosting/
re 3) I'm not sure specifically what happened in your case, but generally servers are provisioned and online in the public cloud within a few minutes -- If I had to guess, its possible the huddle you were in had some kind of capacity issue or other fault that prevented immediate provisioning. If you ever boot a server and its not online in <5 minutes, I'd go straight to support chat, and they generally can tell you what is going on.
She named names, and was given so much grief because of that.
http://markmail.org/message/wyi2r4ntzlmawp6e#query:+page:1+m...
Nothing in cast itself is really linux specific, it should all work on any Unixy-enough OS, and we have plans for how to do it on top of Windows too.
This gets messy because for H1Bs it is difficult to move between jobs, so it is harder for them to get raises, etc, but thats another whole story.
There is no such thing as a well designed and maintained LAN. All networks are unreliable.
I think this is just reinforcing to me to never use a networked filesystem. It is the wrong abstraction for storing data across multiple physical units. Too often you end up hanging the whole box waiting for an IO operation to finish.
NFS fell out of favor for most Linux things a long time ago for these reasons.
IBM's GPFS... I've seen it fail spectacularly. But their marketing materials said it is awesome!
And now Amazon's EBS, essentially a networked block store, is failing.
Don't trust a vendor about networked storage. Its a lie, just like the cake.
Cloudkick (YC Winter 09) was acquired by Rackspace back in December, and we are hiring to fill out our San Francisco office.
Currently looking for Javascript developers and Python/Django developers.
Full listing of openings is here: http://jobs.rackspace.com/search?q=%22san+francisco%22
As part of acquiring Cloudkick (YC W09), Rackspace is building out its first bay area office.
Most of these positions are working with the Cloudkick team.
Looking specifically for:
- Javascript people, who can help build a
fully client side application.
- Python & Node.js people for expanding our
backend services (Cassandra experience is a plus)
- General C/C++/Lua for work on our agent
and monitoring systems.
More specific positions are also up on the racker talent site:
It is called prison.
When a project like Cassandra came to Apache, there were only a couple Facebook developers working on it -- and they had largely abandoned future work on it. After growing the community there are dozens of people today contributing patches from many independent companies -- this makes a much stronger product that will survive the whims of a single company, which is also what consumers of open source projects look for, a stable ecosystem that won't go away if X is bought by Oracle or another BigEvilCorporation.
It remains to be seen if Wave will build a big enough community to be successful inside the ASF, but Cassandra is certainly one of the success stories.
Cloudkick did very well with a freenium model -- it let sysadmins try out our service on their personal machines -- and it led to many using it inside their businesses later.
Those kinds of conversions and upsells are the perfect reason for a B2B application to have a freenium model. It wasn't about getting users for investors, we just looked at it more like a loss-leader and part of our user acquisition costs -- just like the costs of buying Google Ads, etc.
Lots of open positions in San Francisco, most are for parts of the Cloudkick team: http://jobs.rackspace.com/search?q=%22san+francisco%22
Should come by the Cloudkick/Rackspace offices again sometime, would love to catch up.
I just want to respond to one point.
> I can't see RS buying CloudKick for internal
> monitoring of its own customer's servers and
> instances because they would already have a
> mature solution by now.
http://support.microsoft.com/default.aspx?scid=kb;en-us;Q294...
Apache has had a work around for this since.. forever.