Fluentd is a data collector that can unify logging and metrics formats and enable real-time extraction, transformation, and storage of data. It will be used at 10xLab to collect logging data from their Co-Work app and infrastructure components and enable real-time analysis and long-term storage. Fluentd makes it easy to set up log collection pipelines and extend functionality through plugins. 10xLab plans to use Fluentd with Resque to reliably queue and process job data, store logs in S3, analyze logs in Treasure Data, and monitor systems. Fluentd will be installed via AWS cloud-init and managed using Chef.
1 of 61
Downloaded 19 times
More Related Content
Fluentd in Co-Work
1. Fluentd
in
Fluentd meetup in Fukuoka
2013/03/07
@Spring_MT
9. f fluentd
public
subnet
net work
Gateway NAT
SSLの変換
subnet subnet subnet
utility subnet fluentd storage
Co-Work deliver proxy
APP
Watch
f f
redis
f f f
worker varnish
f DB(MySQL on EC2) f +
sinatra
f f f
worker f
f f f
17. worker
worker
app worker
worker
r r queue
e e worker
worker
worker s s
app worker q q
worker
u u queue
e e worker
worker
worker
app worker
worker
18. worker
worker
app
worker
worker
X
If queing fail,
App should retry queuing.
You write a retry
processing?
19. worker
worker
app worker f f
worker l l
u u
e e r queue
n n e worker
worker t t
worker - - s
app worker l r q
worker o e u queue
g s
g q e worker
worker e u
worker r e
app worker
worker
20. worker
app
worker
worker
X X
worker
If queuing fail,
Fluentd retry queuing !
Fluentd(and Logger)
Buffering internally
Retrying