carlivar

116 ms·

541.
by carlivar 11y ago
This could help: http://stackshare.io/erlang OpenX is a big user of Erlang for RTB.
542.
by carlivar 11y ago
Uber is charging a fee which is their costs plus whatever profit they think they can get away with. Minimizing the profit part of the fee sounds admirable.
543.
by carlivar 11y ago
Well this makes me feel better about prior complaints regarding Zookeeper stability at scale. Looks like PagerDuty did a better job troubleshooting!
544.
by carlivar 12y ago
Why?
545.
by carlivar 12y ago
Engineers wanting data isn't exactly surprising. I would not consider it attitude.
546.
by carlivar 12y ago
So you are having ZK downtime? Such a system shouldn't ever really go down, even with Amazon problems. We strive for software in which individual nodes can be flakey or down without impacting the uptime of the whole cluster. ZK doesn&#
547.
by carlivar 12y ago
Yes, I have. Zookeeper's logging (and thus troubleshooting ability) is atrocious.
548.
by carlivar 12y ago
I agree and I want that thing not to be ZK. Consul or etcd would be fine.
549.
by carlivar 12y ago
We would get increased latency interacting with Zookeeper over time until eventually it would completely fail. The log messages when this would happen (either latency or failure) were extremely unhelpful. The logging server-side for Zookeep
550.
by carlivar 12y ago
If only Mesos would drop the Zookeeper requirement. We've found Zookeeper to be overly complicated and difficult to troubleshoot.
551.
by carlivar 12y ago
I appreciate the philosophy of CrunchBang, but for my recently-built Linux desktop PC I tried it out. Spent two days trying to get sound to work correctly. Really brought me back to my struggles with Linux 10+ years ago. Gave up and went wi
552.
by carlivar 12y ago
This is great if you're "all in" on Mesos. However the advantages to a truly independent service discovery system (such as Consul etc.) are that it: * Doesn't tie you to Mesos, as it can move to another "datacenter
553.
by carlivar 12y ago
Yes, I completely agree. Ubuntu is fantastic at auto-detecting your hardware and making everything Just Work (by Linux standards - your mileage may vary). So take advantage of that base but use whatever window manager you want. Easily selec
554.
by carlivar 12y ago
I moved to Linux (Ubuntu) and migrated a HUGE iTunes library along with it. I now use Clementine as my music management / player. I believe there's a way to import an iTunes library, but the iTunes metadata is all in one big XML f
555.
by carlivar 12y ago
Yeah, good questions for sure. I'd suggest asking on the SaltStack IRC channel which is very active and helpful. http://www.saltstack.com/community/
556.
by carlivar 12y ago
How many nodes do you manage? I've heard of Chef scaling horror stories.
557.
by carlivar 12y ago
Salt's REAT protocol uses the CurveCP crypto library, so yes, this is being addressed.
558.
by carlivar 12y ago
We have 2700 machines using a single Salt master. You have to tune it or you have the "thundering herd" problem. There are two parameters if I recall: * a delay between master queries. * randomization of when to check with the mas
559.
by carlivar 12y ago
I feel like every discussion of configuration management should start with what scale you are talking about. Managing 100 servers is quite a bit different than our environment which is about 7500 logical hosts on 5000+ physical servers acro
560.
by carlivar 12y ago
"the problem of orchestrating a bunch of Docker machines is still unsolved." I don't know about that. Mesos was already mentioned: http://mesos.apache.org/ There's also Kubernetes which has recently star
561.
by carlivar 12y ago
I don't think they should be mutually exclusive. etcd is too baked-in to CoreOS. My plan is to try out Consul in containers rather than in CoreOS itself.