Standing Up Your First Cluster
3 likes1,411 views
This document provides guidance on setting up server monitoring, application metrics, log aggregation, time synchronization, replication strategies, and garbage collection for a Cassandra cluster. Key recommendations include: 1. Use monitoring tools like Monit, Munin, Nagios, or OpsCenter to monitor processes, disk usage, and system performance. Aggregate all logs centrally with tools like Splunk, Logstash, or Greylog. 2. Install NTP to synchronize server times which are critical for consistency. 3. Use the NetworkTopologyStrategy replication strategy and avoid SimpleStrategy for production. 4. Avoid shared storage and focus on low latency and high throughput using multiple local disks. 5. Understand
Ad
More Related Content
What's hot (19)
Viewers also liked (7)
Ad
Similar to Standing Up Your First Cluster (20)
Ad
More from DataStax Academy (17)
Recently uploaded (20)
Standing Up Your First Cluster
- 1. ©2013 DataStax Confidential. Do not distribute without consent. Jon Haddad, Technical Evangelist @rustyrazorblade Standing up your first cluster 1
- 3. Server Monitoring & Alerts • Monit • monitor processes • monitor disk usage • send alerts • Munin / collectd • system perf statistics • Nagios / Icinga • Various 3rd party services • OpsCenter - DSE Only • Use whatever works for you
- 4. Application Metrics • Statsd / Graphite • Grafana • Gather constant metrics from your application • Measure anything & everything • Microtimers, counters • Graph events • user signup • error rates • Cassandra Metrics Integration • jmxtrans
- 5. Log Aggregation • Hosted - Splunk, Loggly • OSS - Logstash + Kibana, Greylog • Many more… • For best results all logs should be aggregated here • Oh yeah, and log your errors.
- 6. Get the Server Times Right • Everything is written with a timestamp • Last write wins • Usually supplied by coordinator • Can also be supplied by client • What if your timestamps are wrong because your clocks are off? • Always install ntpd! server time: 10 server time: 20 INSERT real time: 12 DELETE real time: 15 insert:20 delete:10
- 7. Snitches • Snitch lets us distribute data in a fault tolerant way • Changing this with a large cluster is time consuming • Dynamic Snitching • use the fastest replica for reads • DC aware • GossipingPropertyFileSnitch (recommended)
- 8. Replication Strategy • Don't use Simple for prod, you will regret it • NetworkTopologyStrategy = win • Lets you pick replica count per DC
- 9. Don't use Shared Storage • Single point of failure • High latency • Expensive • Performance is about latency • Can increase throughput with more disks • In general avoid SAN, NAS
- 10. Know your compaction! • Compaction merges SSTables • Size Tiered • Just merge stuff, get big • Leveled: SSD • Read Heavy (More I/O) • DateTiered / Timewindow • DT included with OSS • TW written by Jeff Jirsa • Both written to solve Time series • TW easier to configure • DT more options
- 11. Utilities
- 12. nodetool tpstats • What's blocked? • MemtableFlushWriter? - Slow disks! • also leads to GC issues • Dropped mutations? • need repair!
- 13. Proxy Histograms • nodetool proxyhistograms • High level read and write times • Includes network latency
- 14. Table Histograms • nodetool tablehistograms <keyspace> <table> • reports stats for single table on a single node • Used to identify tables with performance problems
- 15. Query Tracing
- 17. JVM GC Overview • What is garbage collection? • Manual vs automatic memory management • Generational garbage collection (ParNew & CMS) • New Generation • Old Generation
- 18. New Generation • New objects are created in the new gen (eden) • Comprised of Eden & 2 survivor spaces (SurvivorRatio) • Space identified by HEAP_NEWSIZE in cassandra-env.sh • Historically limited to 800MB
- 19. Minor GC • Occurs when Eden fills up • Stop the world • Dead objects are removed • Copy current survivor to empty survivor • Live objects are promoted into survivor (S0 & S1) then old gen • Some survivor objects promoted to old gen (MaxTenuringThreshold) • Spillover promoted to old gen • Removing objects is fast, promoting objects is slow
- 20. Old Generation • Objects are promoted to new gen from old gen • Major GC • Mostly concurrent • 2 short stop the world pauses
- 21. Full GC • Occurs when old gen fills up or objects can’t be promoted • Stop the world • Collects all generations • Defragments old gen • These are bad! • Massive pauses
- 22. Workload 1: Write Heavy • Objects promoted: Memtables • New gen too big • Remember: promoting objects is slow! • Huge new gen = potentially a lot of promotion new gen old gen too much promotion
- 23. Workload 2: Read Heavy • Short lived objects being promoted into old gen • Lots of minor GCs • Read heavy workloads on SSD • Results in frequent full GC new gen old gen (full of short lived objects) early promotion fills up quickly
- 24. G1GC • Improvement over ParNew+CMS • Hard to tune • CASSANDRA-8150 • G1 has more predictable pauses • Better latency • Many new gen, many old gen • G1 is adaptive to usage E SO SO E O S EE Eden Old GenS0 S1
- 25. GC Profiling • Opscenter gc stats • Look for correlations between gc spikes and read/write latency • Cassandra GC Logging • Can be activated in cassandra-env.sh • jstat • prints gc activity
- 26. How much does it matter?
- 27. ©2013 DataStax Confidential. Do not distribute without consent. 27