Habitat Features

System & Application Monitor
cpu Real time monitor and visualiser of data metrics from applications and operating system. Built in probes include CPU utilization, storage, memory, network and processes. Take a look at the full list of probes to see what is actually collected.
Graphical Visualisation
Multiple Sources Fast graphical desktop tool displays current data with updates from the local client and other hosts that also run habitat. Data sets can come from anywhere and be retrieved from any time in the past for comparison or later analysis.
Add your own Data

Any user data can be added for time series tracking from either the graphical tool or the command line. Scripting languages of all types can send data to a habitat file or the repository, even in real time! If the data can be made to look like a series of tables, then it can live with habitat (providing its text and numbers!).

Charts or Tables

I/O Charts
Habitat draws line graph charts or text tables of data updated in `near real-time' (depends on underlying data frequency). Multi instance data, like disks, network interfaces or processors are split into adjacent charts.
Drill-down Data

graph rulers
Zoom in to years of data to show a graph of just a few seconds with just a few clicks of the zoom button, then scroll back and forth in time. In Habitat its fast to explore even big time series sets and easy to keep track of with the adaptive time scales
Custom Graphs

Each choice of data set has many potential curves ready to plot. Select them like a menu to plot on the same chart and even scale the big (or small) values to fit in the picture

Storage and Data Routing

Low maintenance local storage All collected data is recorded into a ringbuffer on the machine's local disk, which is a structure that automatically removes old data without the need for administrators. Histories of significant length can be built up over time without resorting to shared archives for absolutely everything, so keeping high performance.

Replication to cental repository The repository is used for to collect data from many machines into one place. It allows long term trends, high density data points and central analysis of data. Its also useful reducing the load on a busy host or looking at a machine's data if it is inaccessible or down. Habitat replicates data to a central archive, provided by the harvest application. It can be read back and treated in the same way as local data.

Multiple data formats Import and export data held in CSV or other tabular formats into habitat's fat headed array format, used to move data around internally. Data can be saved to and read back from local or archived storage with command line tools for batched transfers or interactively with ghabitat.

Data consolidation The secret behind modest storage consumtion is to re-record older data at a lower frequency, thus reducing the number of samples. habitat carries out this operation several times but is still able to reassemble a single continuous set of data. Use the harvest archive for long term high frequency data The parameters for sample frequency, retained quanity and archive replication details can be tuned to suit your circumstances.

Multi-protocol data sources All data sources in habitat are integraded into a common I/O system with a URL-like addresses. Protocols include: file, standard in, out and error, ringstore (the local storage), sqlringstore (the archive), HTTP, HTTPS, even FTP.

More features and programming

User development For developers there is an API for extending the collector (called clockwork) with plug-ins, allowing data to be pulled in from other sources created by users. Data from many sources can then be plotted along side each other.

Working with Grids The combination of habitat and harvest is ideal for working with grids. For small or informal grid infrastructures, habitat on its own is able to connect to hosts directly and browse their data from one point. For larger installations, harvest combines analytics with data archive and organisational information to give clarity to the enterprise.

Log file recording As well as performance data, log files can be captured at regular intervals to provide a context of events. Metrics and events can then be see alongside each other.

Alerts and alarms Patterns in log files or threshold crossing can trigger internal actions and command line execution. This can add an extra level of intellegence to the collection of statistics.

Multi level configuration Highly configurable, habitat allows layers of directives to make maintenance easy. Users can pick up their personal preferences, whilst still using system garden defaults and site customisations.

System & Application Monitor Real time monitor and visualiser of data metrics from applications and operating system. Built in probes include CPU utilization, storage, memory, network and processes. Take a look at the full list of probes to see what is actually collected.	Graphical Visualisation Fast graphical desktop tool displays current data with updates from the local client and other hosts that also run habitat. Data sets can come from anywhere and be retrieved from any time in the past for comparison or later analysis.	Add your own Data Any user data can be added for time series tracking from either the graphical tool or the command line. Scripting languages of all types can send data to a habitat file or the repository, even in real time! If the data can be made to look like a series of tables, then it can live with habitat (providing its text and numbers!).
Charts or Tables Habitat draws line graph charts or text tables of data updated in `near real-time' (depends on underlying data frequency). Multi instance data, like disks, network interfaces or processors are split into adjacent charts.	Drill-down Data Zoom in to years of data to show a graph of just a few seconds with just a few clicks of the zoom button, then scroll back and forth in time. In Habitat its fast to explore even big time series sets and easy to keep track of with the adaptive time scales	Custom Graphs Each choice of data set has many potential curves ready to plot. Select them like a menu to plot on the same chart and even scale the big (or small) values to fit in the picture

Low maintenance local storage	All collected data is recorded into a ringbuffer on the machine's local disk, which is a structure that automatically removes old data without the need for administrators. Histories of significant length can be built up over time without resorting to shared archives for absolutely everything, so keeping high performance.
Replication to cental repository	The repository is used for to collect data from many machines into one place. It allows long term trends, high density data points and central analysis of data. Its also useful reducing the load on a busy host or looking at a machine's data if it is inaccessible or down. Habitat replicates data to a central archive, provided by the harvest application. It can be read back and treated in the same way as local data.
Multiple data formats	Import and export data held in CSV or other tabular formats into habitat's fat headed array format, used to move data around internally. Data can be saved to and read back from local or archived storage with command line tools for batched transfers or interactively with ghabitat.
Data consolidation	The secret behind modest storage consumtion is to re-record older data at a lower frequency, thus reducing the number of samples. habitat carries out this operation several times but is still able to reassemble a single continuous set of data. Use the harvest archive for long term high frequency data The parameters for sample frequency, retained quanity and archive replication details can be tuned to suit your circumstances.
Multi-protocol data sources	All data sources in habitat are integraded into a common I/O system with a URL-like addresses. Protocols include: file, standard in, out and error, ringstore (the local storage), sqlringstore (the archive), HTTP, HTTPS, even FTP.

User development	For developers there is an API for extending the collector (called clockwork) with plug-ins, allowing data to be pulled in from other sources created by users. Data from many sources can then be plotted along side each other.
Working with Grids	The combination of habitat and harvest is ideal for working with grids. For small or informal grid infrastructures, habitat on its own is able to connect to hosts directly and browse their data from one point. For larger installations, harvest combines analytics with data archive and organisational information to give clarity to the enterprise.
Log file recording	As well as performance data, log files can be captured at regular intervals to provide a context of events. Metrics and events can then be see alongside each other.
Alerts and alarms	Patterns in log files or threshold crossing can trigger internal actions and command line execution. This can add an extra level of intellegence to the collection of statistics.
Multi level configuration	Highly configurable, habitat allows layers of directives to make maintenance easy. Users can pick up their personal preferences, whilst still using system garden defaults and site customisations.