Real-time HBase: Lessons from the Cloud

  • View
    1.333

  • Download
    2

  • Category

    Software

Preview:

DESCRIPTION

Speaker: Bryan Beaudreault (HubSpot) Running HBase in real time in the cloud provides an interesting and ever-changing set of challenges -- instance types are not ideal, neighbors can degrade your performance, and instances can randomly die in unanticipated ways. This talk will cover what HubSpot has learned about running in production on Amazon EC2, how it handle DR and redundancy, and the tooling the team has found to be the most helpful.

Citation preview

Real-time HBase: Lessons from the Cloud

Bryan Beaudreault, @HubSpotDev

You’re doing it WRONG!

Instance types

Network,Neighbors,

PERSPECTIVE:

Consumer/Prospect

PERSPECTIVE:

Consumer/Prospect

PERSPECTIVE:

Consumer/Prospect

PERSPECTIVE:

Consumer/Prospect

Improve reads, limit impact

PERSPECTIVE: Consumer/

Prospect

PERSPECTIVE: Consumer/

Prospect

PERSPECTIVE: Consumer/

Prospect

Over-provision, fail fast

\PERSPECTIVE:

Consumer/Prospect

PERSPECTIVE: Consumer/

Prospect

PERSPECTIVE: Consumer/

Prospect

PERSPECTIVE: Consumer/

Prospect

PERSPECTIVE: Consumer/

Prospect

PERSPECTIVE: Consumer/

Prospect

PERSPECTIVE: Consumer/

Prospect

c1.xlarge

i2.4xlarge

CPU heavy workloadsReduce memory footprintAdd more servers

Excellent, but expensiveUse data encoding to reduce diskUse Java7 and G1 to reduce GCs

m1.xlarge Memory heavy workloadsBeware compactionsMore regions

PERSPECTIVE: Consumer/

Prospect

cpu fsWriteLatency

fsReadLatencyload

callQueueLen

compactionQueueSizeiowait

steal

heap

locality

requests

memstoreSizeMB

blockCacheHitCachingRatio

flushQueueSize

PERSPECTIVE: Consumer/

Prospect

Master HBase with us.

dev.hubspot.com

Bryan Beaudreault @HubSpotDev

Recommended