Ready to know about downtime before your customers?
Status List delivers uptime monitoring and professional hosted status pages for sites of all shapes and sizes.
Trusted by 1000+ companies
ZooKeeper Monitoring Guide
-
How do I monitor ZooKeeper?
We can use platform integrations, self hosted options or third party providers. Platform integrated solutions make devops really simple, but might not have all the detail you would like. Self hosted solutions can be highly customizable but require complex setup and ongoing maintenance. Third party solutions can be great if you find the right partner. Let's dive into some specifics.
Platform Monitoring Solutions
Platforms have some great built-in monitoring and alerting tools. Consider plugging ZooKeeper into your platform’s tooling. It will save you a lot of time reinventing the wheel. A platform solution will save you from hunting in different places for your monitors.
Build Your Own Monitoring (self hosted)
Build your own monitoring is a good solution for companies who want high customization and control and don’t mind the operational overhead. Self hosted monitoring tools for ZooKeeper offer a wide selection of metric databases and graphing libraries. You can piece together your favorites and have a great system. You also have full control over upgrades and changes to your monitoring system. No more worrying about a 3rd party making unexpected changes. But, this comes at a cost.
Self hosted monitoring does require some operation overhead. You need to maintain the servers and monitoring software that monitor your ZooKeeper cluster. Your team will need to be on call to fix operational issues if they arise. If operations is something your company excels at, this may be the solution for you. If you don’t have an operations team, or tend to outsource most of your solutions, this may not be the best solution for you.
3rd Party Monitoring Solutions
Third party providers have some great tools without the devops overhead. You get a highly specialized tool and 24/7 support with very little setup work. If you aren’t sure which solution is right for you, this is probably the one for you.
1. DataDog
Pricing: 15/host/mo
Very easy to use.
- Thread & JVM usage,
- Latency, Znode Count etc
- Leader change
Installation: On host: Update configuration file, restart agent
Click image to expand
2. Librato (SolarWinds/AppOptics)
Pricing: 9.99/host/mo (buy in blocks of 10 hosts)
Very easy to use.
- Thread & JVM usage,
- Latency, Znode Count etc
- Missing Leader Change
Installation: Update configuration file, restart agent.
Click image to expand
3. SemaText
Pricing: 10.08/host/mo
Easy to use.
- Thread & JVM usage (separate plugin)
- Latency, Znode Count etc (mntr output)
- Missing Leader Change
Installation: Update configuration file, restart agent.
Click image to expand
4. Site24x7 (ManageEngine)
Pricing: 9.00/mo (buy in blocks of 10 hosts)
Decently easy to use
- Thread & JVM usage
- Latency, Znode Count etc
- Missing Leader Change
Installation: Clone python file, update configuration file, restart agent.
Integration Details | Pricing Details (infrastructure tab)
Click image to expand
5. Splunk
Pricing: Call for quote
Somewhat difficult to use
- Thread & JVM usage
- Latency, Znode Count etc
- Leader Change
Installation: Update configuration file, restart agent.
Click image to expand
6. New Relic
Pricing: 0.30/GB ingested + $49/user/mo
Somewhat difficult to use, integration is a community project
- Thread & JVM usage (separate plugin)
- Latency, Znode Count etc
- Missing Leader Change
Installation: Download special binary, update config, permissions, restart agent
Click image to expand
7. LogicMonitor
Pricing: Request quote
Somewhat difficult to use
- Thread & JVM usage
- Latency, Znode Count etc
- Leader Change
Installation: Unknown
Click image to expand
8. WaveFront (VMWare)
Pricing: Request quote
Somewhat difficult to use
- Thread & JVM usage (unclear if supported)
- Latency, Znode Count etc (unclear if supported)
- Leader Change (unclear if supported)
Installation: Update configuration file, restart agent
Click image to expand
9. ManageEngine (paid, on-prem)
Pricing: Call for quote
Somewhat difficult to use.
- Thread & JVM usage
- Latency, Znode Count etc (JMX)
- Leader Change
Installation: Open GUI, set JMX configuration
Click image to expand
9. ITRSGroup (paid, on-prem)
Pricing: Sign up for demo
Somewhat difficult to use.
- Thread & JVM usage
- Latency, Znode Count etc (JMX)
- Missing Leader Change
Installation: Download zip file, run package installer, update configuration
Click image to expand
10. Instana (IBM)
Pricing: 75/host/mo (buy in blocks of 10 hosts)
Somewhat difficult to use.
- Thread & JVM usage
- Latency, Znode Count etc (missing a lot of metrics)
- Missing Leader Change
Installation: Update configuration file, restart agent
Click image to expand
11. Fortinet (paid, on-prem)
Pricing: Call for quote
Somewhat difficult to use.
- Thread & JVM usage
- Latency, Znode Count etc
- Missing Leader Change
Installation: Update configuration file, restart agent
Click image to expand
12. Sensu (paid, on-prem)
Pricing: 3.00/host/mo
Somewhat difficult to use. Takes several open source softwares and bundles them together as a package.
- Thread & JVM usage
- Latency, Znode Count etc (mntr output)
- Missing Leader Change
Installation: Run CLI command to add plugin
Click image to expand
12. Zabbix (open source, on-prem)
Pricing: Call for quote (for professional support)
Somewhat difficult to use.
- Thread & JVM usage
- Latency, Znode Count etc (mntr output)
- Leader Change
Installation: Add yaml file, restart agent
Click image to expand
Ready to know about downtime before your customers?
Status List delivers uptime monitoring and professional hosted status pages for sites of all shapes and sizes.
Trusted by 1000+ companies