github DataDog/dd-agent 5.8.0
on GitHub

latest releases: 5.32.9, 5.32.8, 5.32.8-rc.7...
7 years ago

5.8.0 / 05-23-2016

All platforms

Details

5.7.4...5.8.0

New integrations

  • Kong
  • PowerDNS Recursor
  • Spark

Updated integrations

  • ActiveMQ
  • Cassandra
  • Directory
  • Disk
  • Docker
  • Elasticsearch
  • Ganglia
  • Graphite
  • HAProxy
  • HTTP Check
  • JMX
  • Kubernetes
  • MapReduce
  • Marathon
  • Memcached
  • Mesos Master
  • MongoDB
  • MySQL
  • Process
  • RabbitMQ
  • Riak
  • TeamCity
  • TokuMX
  • Varnish
  • vSphere
  • Windows
  • Windows Service
  • WMI
  • YARN

Service Discovery

Starting with the 5.8 release, the Datadog Agent ships with service discovery. This feature works for Docker containers and can run on platforms such as Kubernetes, Docker Swarm, and Amazon ECS.

Service discovery makes it possible to define configuration templates for a given Docker image in a configuration store, and to dynamically apply them to new containers on the fly. The Agent will then use these configuration templates combined with container metadata to enable, disable and reconfigure checks dynamically in response to container start/stop events.

For more details on Datadog Agent's Service Discovery, check out the wiki page.

See #2008, #2424, #2426, #2514.

Kong integration

Kong is an open-source management layer for APIs.

The checks collects metrics using /status endpoint part of Kong RESTful API. It retrieves usage information about the node, with some basic information about the connections being processed by the underlying NGINX process, and the number of entities stored in the datastore collections.
It also submits a service check to reflect the agent's ability to connect to the instance.

See #2241.

A big thank you @shashiranjan84 !

Spark integration

The Spark check runs from the Master Node and retrieves metrics from YARN's ResourceManager.
The collected metrics cover statistics about:

  • Spark Jobs
  • Spark Stages
  • Spark Executors
  • Spark Resilient Distributed Datasets (RDD)

See #2407.

PowerDNS Recursor integration

The PowerDNS Recursor check retrieves metrics using the /servers/localhost/statistics endpoint part of PowerDNS API when the experimental web server is active.
The collected metrics cover the Recursor statistics: (outgoing) queries, answers, cache and cache entries.
A service check reflects the agent's ability to connect to the instance.

For more details about the metrics collected, check out the PowerDNS documentation about Recursor Statistics.

See #2108, #2490.

A big thank you @AntoCard, @janeczku !

MapReduce & YARN updates

Hadoop MapReduce and YARN integrations have been updated to reduce and consolidate the metrics' tags.
In particular:

  • cluster_name tag replaces cluster_id
  • app_name tag replaces app_id

MapReduce metrics have received additional improvements and cleanups:

  • Metric type rate replaces gauge when applicable
  • The following metrics have been removed:
    • mapreduce.job.map.task.progress
    • mapreduce.job.reduce.task.progress

See #2473, #2474, #2487.

Source install: end of Python 2.6 compatibility support

For source installation, the Datadog Agent uses system's Python. Starting with the 5.8.0 release, the Datadog Agent does not enforce compatibility with Python 2.6 anymore and requires Python 2.7. Tests and code sections used to maintain compability have been removed.

Users running a different version of Python on their system are advised to install a packaged version of the Datadog Agent. If you have any questions, please contact our support team support(at)datadoghq(dot)com.

See #2377.

Changes

  • [FEATURE] ActiveMQ: Add monitoring for standby hosts. See #2023, #2430 (Thanks @joelvanvelden)
  • [FEATURE] Directory: Add Windows support. See #2410, #2508
  • [FEATURE] Disk: Add an option to exclude disks based on their mountpoints. See #2359, #2476 (Thanks @rhwlo)
  • [FEATURE] Docker: Add optional disk metrics. See #2405
  • [FEATURE] Docker: Add the possibility to tag by container ID. See #2353, #2507
  • [FEATURE] Elasticsearch: Add support for SSL and Elasticsearch Shield. See #2195 (Thanks @pabrahamsson)
  • [FEATURE] Elasticsearch: Collect additional Thread Pool metrics. See #2321
  • [FEATURE] HAProxy: Add option to disable SSL validation. See #2393.
  • [FEATURE] JMX: Add user tags to service checks. See #96
  • [FEATURE] JMX: Allow group name substitutions in attribute/alias parameters. See #94, #97 (Thanks @alz)
  • [FEATURE] Kong: New check. See #2241, Kong integration (Thanks @shashiranjan84)
  • [FEATURE] Kubernetes: Allow to blacklist labels. See #2512
  • [FEATURE] Kubernetes: Tag pods metrics with replica sets. See #2444, #2446, #2512
  • [FEATURE] Memcached: Add "items" and "slabs" statistic metrics as reported by the stats command. See #2491
  • [FEATURE] PowerDNS Recursor: New check. See #2108, #2490, PowerDNS Recursor integration (Thanks @AntoCard, @janeczku)
  • [FEATURE] Process: Add pagefault statistics to process metrics. See #2363, #2477 (Thanks @ovesh)
  • [FEATURE] RabbitMQ: Tag queue based metrics with the queue name as queue_family. See #2259, #2468 (Thanks @rhwlo)
  • [FEATURE] Riak: Collect Riak Search metrics. See #2243 (Thanks @gzysk8)
  • [FEATURE] Service Discovery: Define configuration templates for a given Docker image in a configuration store, and to dynamically apply them to new containers on the fly. See #2008, #2424, #2426, #2514, Service Discovery
  • [FEATURE] Spark: New check. See #2407, #2479, Spark integration
  • [FEATURE] TeamCity: Support basic HTTP authentication. See #2266 (Thanks @mderomph-coolblue)
  • [FEATURE] vSphere: Add the ability to set a custom field from the vSphere PowerCLI to mark the VMs to monitor. See #2459
  • [FEATURE] Windows Event Log: Add an optional event_format parameter to instruct the check to generate Datadog's event bodies with the specified list of event properties. See #2441
  • [FEATURE] Windows: Add opened handles statistics to system metrics. See #2403
  • [FEATURE] Windows: Add swap -pagefile- statistics to system metrics. See #2422
  • [IMPROVEMENT] Core: Allow AgentCheckTest facilities to be used on SDK check modules. See #2341
  • [IMPROVEMENT] Core: Gracefully handle responses with HTTP status code 413. See #2418
  • [IMPROVEMENT] Core: Update the configuration logic to look for checks in a 3rd-party directory. See #2273, #2467
  • [IMPROVEMENT] Directory: Speed up the file count in a directory. See #2245 (Thanks @tebriel)
  • [IMPROVEMENT] Gohai: Collect process resources and deprecate the corresponding Python check. See #24, #2218
  • [IMPROVEMENT] Gohai: Improve logging. See #30
  • [IMPROVEMENT] Kubernetes: Enable kubelet check by default. See #2512
  • [IMPROVEMENT] Kubernetes: Improve support for Kubernetes >= 1.2. See #2388, #2494, #2512
  • [IMPROVEMENT] Kubernetes: Timeout on slow requests. See #2345
  • [IMPROVEMENT] MapReduce: Consolidate and reduce the number of tags produced. See #2474, MapReduce & YARN updates
  • [IMPROVEMENT] MapReduce: Switch metric types from gauge to rate when applicable. See #2487, MapReduce & YARN updates
  • [IMPROVEMENT] Marathon: Remove versions metric to decrease the check collection runtime. See #1861, #2443 (Thanks @Zarkantho)
  • [IMPROVEMENT] MongoDB: Add additional tag on status change to make searching more precise. See #2502 (Thanks @gphat)
  • [IMPROVEMENT] MongoDB: Improve metric tagging. See #2457
  • [IMPROVEMENT] MySQL: Improve accuracy for query run time averages. See #2406, #2463
  • [IMPROVEMENT] YARN: Consolidate and reduce the number of tags produced. See #2473, MapReduce & YARN updates
  • [BUGFIX] Core: Fix DogStatsd's service check parsing when a tag key ends with m. See #2373, #2472
  • [BUGFIX] Core: Restart the forwarder when a "pathological" activity is detected. See #2322
  • [BUGFIX] Core: Retrieve EC2 instance ID on Windows. See #2451
  • [BUGFIX] Docker: Fix container ID extraction for RancherOS. See #2454, #2466 (Thanks @ricky26)
  • [BUGFIX] Docker: Fix improper logging causing the check to raise an exception when excluding containers. See #2464 (Thanks @yenif)
  • [BUGFIX] Docker: Ignore mountpoints not accessible by Datadog Agent dd-agent user. See #2427 (Thanks @c960657)
  • [BUGFIX] Elasticsearch: Restrict the list of metrics to collect by version. See #2186, #2232 (Thanks @bdharrington7)
  • [BUGFIX] HTTP Check: Support no_proxy environment variable. See #2448
  • [BUGFIX] JMX: Report properly beans with ':' in the name. See #90, #91, #95 (Thanks @bluestix)
  • [BUGFIX] JMX: Sanitize metric names and tags, i.e. remove illegal characters. See #89
  • [BUGFIX] JMX: Support javax.management.Attribute attribute types. See #92 (Thanks @nwillems)
  • [BUGFIX] JMX: Use localhost interface when StatsD is bound to all interfaces. See #2516
  • [BUGFIX] Kubernetes: Handle case where no pods are running. See #2263, #2512
  • [BUGFIX] Mesos Master: Fix key error when a metric is missing. See #2500 (Thanks @gphat)
  • [BUGFIX] MongoDB: Attach the check's events with the integration in Datadog. See #2456
  • [BUGFIX] Packaging: Exclude an operating-system-specific .dll from the MSI installer that was causing installation and runtime failures on Windows Server 2008. See #2486
  • [BUGFIX] TokuMX: Fix AttributeError exceptions when collecting metrics from a replica. See #2390
  • [BUGFIX] TokuMX: Fix an issue among metrics collection causing rate metric values to be potentially wrong and ZeroDivisionError exceptions. See #2510
  • [BUGFIX] TokuMX: Support metric values with Int64 type. See #2390
  • [BUGFIX] Windows Service: Make Windows service names case-insensitive. See #2492
  • [OTHER] Cassandra: Remove deprecated options from the datadog.conf configuration file. See #2447
  • [OTHER] Core: Remove support of Python 2.6. See #2377, Source install: end of Python 2.6 compatibility support
  • [OTHER] Core: Upgrade boto dependency from 2.36.0 to 2.39.0. See #2402, #3399
  • [OTHER] Disk: Remove deprecated options from the datadog.conf configuration file. See #2447
  • [OTHER] Ganglia: Remove deprecated options from the datadog.conf configuration file. See #2447
  • [OTHER] Graphite: Remove deprecated options from the datadog.conf configuration file. See #2447
  • [OTHER] HTTP Check: Remove the deprecated threshold and window parameters from the configuration file. See #2400
  • [OTHER] Memcached: Document usage for listening to a Unix socket. See #2298, #2431 (Thanks @kzw)
  • [OTHER] TokuMX: Document usage for connecting to a MongoDB instance with a "full" URI. See #2390
  • [OTHER] Varnish: Document varnish group requirement. See #2384
  • [OTHER] WMI: Document wildcard usage in filters. See #2368, #2469 (Thanks @mausch)
Check out latest releases or
releases around DataDog/dd-agent 5.8.0

Don't miss a new dd-agent release

NewReleases is sending notifications on new releases.

Get notifications