Since our initial launch in mid-2020, GridGain Control Center has strived to bring transparency and flexibility to the monitoring and development of Apache Ignite and GridGain applications. With each monthly update, we introduce new features to make it easier for admins and developers to understand what exactly is happening within their clusters. In our latest update (2020.11.00), we add new features for Cache monitoring, new API instrumentation for Tracing, and new security integrations.
I will be doing a training session on how to manage and monitor Ignite with Control Center on January 20, 2021 at 10am PT. Please join the session to get a comprehensive, hands-on experience with the latest version of Control Center.
Caches View
While it has always been possible to build custom dashboards that display important Cache metrics, it required knowledge of which metrics to choose and how best to display them. The Cache view is a new top level feature which collects these metrics in a simpler layout and adds a new set of actions. With the Cache view, you can view the list of caches in the cluster, understand data partition distribution across the nodes, and track lost partitions in the event of a problem with the cluster. New actions have also been added for cache management including the ability to initiate a rebalance on a particular cache and options to clear data and delete caches from the cluster.
Tracing Improvements
To assist in debugging and performance analysis, Gridgain (and more recently Apache Ignite) introduced application tracing based on OpenCensus. As distributed applications scale horizontally to meet their performance needs, it can be tricky to track API calls as they traverse the nodes of the cluster. Control Center allows you to collect the individual API calls as span data and then combine the spans into a unified call stack, or trace.
New APIs have recently been instrumented to support tracing analysis including Transactions, Cache Read, and Cache Write operations. Being able to analyze the timing of individual steps in a transaction makes it easier to pinpoint bottlenecks in the execution stack. While this won’t fix an issue, it will give you a good idea of where to start looking during root cause analysis.
For a video demonstration of how to leverage tracing to identify bottlenecks in distributed operations running on a Ignite cluster, watch this video.
Control Center has also made it more convenient to enable and disable tracing by simply toggling data collecting for the API of interest. This gives you the power and flexibility to decide when you need to get that extra bit of debugging information and then turn off tracing collection as needed.
Single Sign On Integration
To make integration easier with on premise installations, Control Center added integration for OpenID Connect to make it easier to work with environments that use Single Sign On infrastructure. Through OpenID Connect, Control Center can support common identity providers such as; Okta, Google, Microsoft/Skype, and more.
For a short tutorial on how to connect Control Center to your custom OpenID provider, check out this video. For GridGain EE and UE users, watch this video to see how to OpenID Connect SSO with both GridGain and Control Center.
What’s Next
Control Center continues to march on adding new features to simplify development for Apache Ignite and GridGain. In the coming months, you will see new ways to search and learn about the hundreds of metrics exposed from your clusters. You will have more options to organize metrics dashboards into reusable templates with sample templates focused on different topics like CPU, memory utilization, transactions, threading, and more. Also you will also see new options to integrate into existing infrastructure like LDAP and Active Directory.
For additional information about GirdGain Control Center check out these resources.
- Online Training - How Monitor and Manage Apache Ignite With GridGain Control Center: Hands-on training session for those wondering how to monitor and manage Apache Ignite clusters in production on January 20, 2021 at 10am PT.
- Introduction to Control Center for Apache Ignite: A deep dive into the various capabilities of Control written by Denis Mekhanikov whose team developed Control Center.
- Connecting Ignite clusters to Control Center: Two part video shows how to connect an Ignite cluster to Control Center (part 1) and how to update the application configuration to leverage Control Center (part 2)
- Leverage Tracing to Troubleshoot Ignite Performance Issues: Video demo of how to identify bottlenecks in distributed operations, running over an Apache Ignite cluster, with OpenCensus and GridGain Control Center tracing capabilities.
The Control Center software is available as a hosted version or as a downloadable version. To get started sign up for the hosted version or download the on-prem version.