Embedded Files
What does Uhoh do?
What does Uhoh do?
To keep your digital production line running smoothly, you need to know what's happening. And, as IT ecosystems become ever more complex and distributed in nature, keeping a close eye on everything becomes more and more difficult. So to stay on top of this problem, you need tools which can worm their way into your environment, scooping up every little morsel of information, analysing it, and presenting it in an easily digestible style - a style that makes sense for your business objectives.
Welome to Uhoh.
Uhoh is a super-lightweight, flexible and easy to use distributed system monitoring toolkit. Packed with a plethora of neat little features for slurping up log files and collecting streams of information from your applications, Uhoh binds together your End-to-End Service Assurance stack to ensure that the flow of essential events and metrics can be put to good use.
What do we mean by "Service Assurance" ?
What do we mean by "Service Assurance" ?
There are many facets to Service Assurance, but from the point-of-view of Uhoh we are talking about:
- The ability to map the events from the technical monitoring of elements through to impact on services or stages of a customer journey.
- The ability to measure, over a period of time, whether a service is operating with a designed or contracted level of availability.
In other words, from a business perspective, to be able to answer the question - "is it working ?".
Coupling Uhoh's data collection and processing framework with it's service-mapping capabilities, delivers a really solid base for fully-automated service assurance.
How does Uhoh work?
How does Uhoh work?
Deploying Uhoh into your environment involves:
- Running one or more central Uhoh Server instances.
- Running an Uhoh Client instance on each host which needs to be monitored.
- Configuring Uhoh to determine what it should monitor and what it should do with captured events.
The Uhoh Server's main functions are to:
- Collect alerts and metrics raised by Uhoh Clients and write them to a Log Stream.
- Provide:
- Fault Management views showing currently-active alerts.
- Performance Management views showing recorded metrics.
- Service Map views depicting the live status of service layers and components within your system.
- REST/JSON APIs which provide fault and performance management data to external applications.
- Supply Uhoh Clients with their monitoring configuration. Configuration is always fetched from Uhoh Servers, not held locally on the hosts running Uhoh Clients.
- Advertise the Uhoh Server's presence to Uhoh Clients so that Clients know where to fetch their monitoring configurations from and where to forward alert and performance metric data to.
- Monitor for Uhoh Clients which have closed down unexpectedly (or more likely that the hosts that they are running on have closed down).
The Uhoh Client running on a host:
- Consumes log files, searching for specific patterns and transforming log information into normalised formats.
- Aggregates log information, generating metrics through simple calculations.
- Correlates events and generates further alerts.
- Checks that specific processes are running.
- Monitors host metrics (CPU, disk, memory etc.)
- Consumes alerts from instrumented applications via UDP or web-hooks.
- Checks that network services such as APIs are available.
- Forwards alerts and metrics to Uhoh Servers.
The main focus for Uhoh is to provide the often missing 'glue' in between applications and analytics or visualisation tools. For example an Uhoh Server Log Stream can be used by simple scripts which feed quality-controlled alert and metric data into ElasticSearch, Amazon Web Services CloudWatch, a Slack channel used by your DevOps teams or any other metric or alert visualisation tool.
Uhoh doesn't require any agents or implants to be embedded within your applications.
Uhoh Clients feed events to Uhoh Servers which in turn feed events to Uhoh Log Streams.
The Uhoh Server Fault Management View.
How do I use Uhoh?
How do I use Uhoh?
At the most basic level, using Uhoh involves setting up an Uhoh Server and then running Uhoh Clients on all of your hosts. Full details of how to set up and Uhoh Server can be found on this page: Set up a Server. A single Uhoh Server provides you with the full range of fault and performance management capabilities offered by Uhoh, but for real-world deployments, you will most likely require:
- Multiple Uhoh Servers for resilience.
- Chaining of Uhoh Servers in order to partition monitoring coverage over multiple domains.
- Integration into other systems via the Uhoh Server Log Stream and REST/JSON APIs.
The Bring it to Life page covers all of the above and more.
Both the Uhoh Server and Uhoh Client are implemented 100% in Java and are not dependent on any external Java libraries. They are both supplied within a single Java archive (.jar) file and run standalone rather than needing to be deployed into any kind of container. The various user interfaces (Fault Management, Performance Management, Service Map View and Dashboard) make use of HTML and Javascript. Javascript libraries and fonts are loaded from Internet-facing (Mootools, Google Charts and Google Fonts) sources as needed. Uhoh was designed to run on Linux / Unix-like systems - requiring only a Java Runtime of version 1.6 or above. Also, as Uhoh is so teeny-tiny, it's very, very easy to test new monitoring configurations on a single host.
For a deep-dive into how to configure monitoring functionality within Uhoh, refer to the Set up a Client page.
Where can I get Uhoh?
Where can I get Uhoh?
Uhoh is released as Open Source software under the FreeBSD licence. It can be downloaded from GitHub here.
Google Sites
Report abuse