Release notes#

Learn about new features, enhancements, known issues, resolved issues, and other important details for Rackspace Monitoring API 1 service in the following release note information.

Note

For information about using the API, see the documentation overview.

v1.20, January 31, 2016#

These release notes correspond to a new-feature release of Rackspace Monitoring.

What’s new#

This release of the monitoring agent includes OpenSSL 1.0.2e.

Resolved issues#

The following agent-specific changes were made:

  • Plugins: Executables within directories of the whitelisted folder were not being ran correctly

  • ‘-N’ command line argument: Skips entity creation on setup

  • Set the default locale

  • Write the configuration file slightly different to see if extraneous characters on some machines gets fixed

  • Fixed an issue in an underlying library performing resolv.conf parsing. If there was more than one whitespace character in-between the nameserver and ip then the application would not discover the DNS servers correctly. This release supports the use of these extra characters.

Known issues#

None for this release.

v1.19, December 31, 2015#

These release notes correspond to a new-feature release of Rackspace Monitoring.

What’s new#

This release of the monitoring agent includes OpenSSL 1.0.2e.

Resolved issues#

The following agent-specific changes were made:

  • Made a fix for the Apache status page check to work correctly with virtual host based systems

  • Made a fix for Windows where an underlying library was returning a NULL.

  • Fixed a crash and adds more logging to invalid responses from the server endpoint.

Known issues#

None for this release.

v1.18, November 30, 2015#

These release notes correspond to a new-feature release of Rackspace Monitoring.

What’s new#

Luvi was upgraded to 2.3.5.

Resolved issues#

The following agent-specific changes were made:

  • On Windows, made a change to properly close stdin when gathering the computer ID.

  • Made various hostinfo improvements.

  • Added a fix for an issue with gathering all available metrics within custom plug-ins.

Known issues#

None for this release.

v1.17, October 22, 2015#

These release notes correspond to a new-feature release of Rackspace Monitoring.

What’s new#

The agent for Ubuntu 15.10 was released.

Resolved issues#

The following agent-specific changes were made:

  • Added CPU count to Windows PerfOS check

  • Made a change to close stdin before running a check, which fixes a hang on certain flavors of Windows with custom plug-ins.

  • For Red Hat Enterprise Linux 7, fixed the post install script for detecting chkconfig.

  • Made an update to propagate hostinfo only for the current running operating system.

Known issues#

None for this release.

v1.16, September 18, 2015#

These release notes correspond to a new-feature release of Rackspace Monitoring.

What’s new#

Several hostinfo additions and changes were made. For more information, see Hostinfochecks.

Resolved issues#

The following agent-specific changes were made:

  • Added a fix for the metrics time stamp.

  • Added a fix to propagate the correct GUID within the protocol exchange.

  • Upgraded libuv to 1.7.4.

Known issues#

None for this release.

v1.15, August 15, 2015#

These release notes correspond to a new-feature release of Rackspace Monitoring.

What’s new#

None for this release.

Resolved issues#

The following agent-specific changes were made:

  • Fixed a bug in the socket timeout logic found for service net monitoring agents (including DBaaS).

  • Fixed a bug on service net reconnects.

  • Improved the hostinfo infrastructure for future releases.

  • Updated various components within Luvit for improved stability.

  • Added fixes for hostinfo and disabling prelink on RPM systems.

  • Added a fix for the RPM packaging for the previous 2.1.16 release.

  • Fixed an issue with Ubuntu 15.04 and systemd registration across reboots.

  • Upgraded an underlying library (libuv) that the agent uses for asynchronous I/O operations. There was a regression in this library that could potentially cause performance issues with file system operations.

Known issues#

None for this release.

v1.14, July 29, 2015#

These release notes correspond to a new-feature release of Rackspace Monitoring.

What’s new#

  • Automatic upgrades on Linux and Windows were performed.

  • New Luvi and Luvit infrastructure was added.

  • A number of new hostinfo modules were added: pam, fstab, cron, login, fileperms, kernel_modules, and services. For more information, see Hostinfo checks.

Resolved issues#

The following agent-specific changes were made:

  • Fixed OpenSSL memory leaks.

  • Significantly improved the process for creating agents.

  • Made a change to use a built-in libsigar library (when /tmp is mounted with the noexec option).

  • Use a built-in libsigar library (when /tmp is mounted with the noexec option).

  • Upgraded OpenSSL to 1.0.2d.

  • Added a gethostname fix for Windows.

  • Added a fix for the startup sequence for Windows.

  • Added a bug fix for zombie processes with agent plug-ins on UNIX platforms.

  • Fixed a minor issue with enumerating the new hostinfo modules via raxmon-agent-host-info-types.

  • Fixed a bug found in unit tests with a garbage collect in the uv_spawn wrapper.

Known issues#

None for this release.

v1.13, May 27, 2015#

These release notes correspond to a new-feature release of Rackspace Monitoring.

What’s new#

Support was added for Fedora 22.

Resolved issues#

The following agent-specific change was made:

Added a fix for systemd networking dependency (notably on Red Hat Enterprise Linux 7 and CentOS 7 flavors)

Known issues#

None for this release.

v1.12, April 26, 2015#

These release notes correspond to a new-feature release of Rackspace Monitoring.

What’s new#

  • Support for Debian 8 (Jessie) was added.

Resolved issues#

The following agent-specific changes were made:

  • Added a small patch to the CPU check for memory usage.

  • Added a small patch to support machine enumeration on newer Windows instances that lack the xenstore.exe binary.

  • Trimmed the Monitoring ID and Monitoring Token configuration variables to ensure correctness.

  • Fixed the Xen Server 6 package repository.

Known issues#

None for this release.

v1.11, March 28, 2015#

These release notes correspond to a new-feature release of Rackspace Monitoring.

What’s new#

List checks

for an entity.

Resolved issues#

The following agent-specific changes were made:

  • Added support for Fedora 22.

  • Fixed * systemd networking dependency (notably on Red Hat Enterprise Linux 7 and CentOs7).

  • Added support for Debian 8 (Jessie).

  • Trimmed the Monitoring ID and Monitoring Token configuration variables to ensure correctness.

  • Fixed the Xen Server 6 package repository.

  • Implemented support for machine enumeration on newer Windows instances that lack the xenstore.exe binary.

  • Modified check scheduling to prevent a thundering herd issue.

  • Changed garbage collection to incrementally garbage collect.

  • Optimized a few parser components to use less memory.

  • Fixed a thundering herd issue for test checks.

  • Tuned OpenSSL options (compression and buffers).

  • Fixed memory leak with new connections.

  • Fixed an issue with the MySQL check on DBaaS (and other MySQL instances) of not closing the connection gracefully.

  • Upgraded OpenSSL to 1.0.1L.

  • Implemented a fix for automatic upgrade logging after an update.

Known issues#

None for this release.

v1.10, January 23, 2015#

These release notes correspond to a new-feature release of Rackspace Monitoring.

What’s new#

Resolved issues#

This release includes the following agent-specific changes:

  • Added a new metric on Linux to show whether a networking link is up or down.

  • Implemented a fix for automatic upgrades.

  • Fixed a memory leak for Luvit.

  • Fixed a connection memory leak.

  • Added support for Fedora 21.

Known issues#

None for this release.

v1.9, December 1, 2014#

These release notes correspond to a new-feature release of Rackspace Monitoring.

What’s new#

  • Email notifications were improved to indicate when an observation was too old to be included in a consistency calculation.

  • In-place automatic agent upgrades was released and documented. Starting with agent version 1.1.0-15, all installed agents will automatically upgrade when a new version is released. You can opt out of automatic upgrades. For details, see the Developer Guide <https://developer.rackspace.com/docs/cloud-monitoring/v1/developer-guide/>__

  • The MySQL agent check type was released and documented. For details, see the Developer Guide.

Resolved issues#

The following agent-specific changes were made:

  • Fixed a memory issue in an upstream package.

  • Corrected behavior for configuring properties with ServiceNet.

  • Fixed log rotation permissions for Red Hat based systems.

  • Added endpoint to return targets for agent plug-ins.

  • Added endpoint to return the list of plug-ins.

  • Fixed memory leak in FS_CALL.

  • Implemented fix for custom endpoints.

  • Implemented fix for PowerShell parameter expansion.

  • Fixed issue with the MySQL check type.

  • Fixed issue with capturing a socket error on connect

  • Added Windows MSI installation for all users by default. This fixes some issues with the Rackspace Monitoring Agent not appearing in the Windows Add and Remove programs feature when installed by one user and viewed by another.

  • Implemented change to not close a MySQL connection on query errors.

  • Added support for HTTP proxy CONNECT.

  • Upgraded MySQL check with new metrics.

  • Added several minor bug fixes and reliability improvements.

Known issues#

None for this release.

v1.8, August 13, 2014#

These release notes correspond to a new-feature release of Rackspace Monitoring.

What’s new#

  • An SMS notification type is now available. This notification type enables Rackspace Monitoring users to receive notifications on an SMS-enabled mobile phone.

This API-only feature is available globally without regard to region.

This API-only feature is available globally without regard to region.

  • Service level names have changed:

    • The Infrastructure service level has changed to the Managed Infrastructure service level.

    • The Managed Cloud service level has changed to the Managed Operations service level.

Resolved issues#

  • Fixed a crash on Linux if the PID file could not be read.

  • Upgraded OpenSSL to 1.0.1i.

  • Added an agent host_info_type message to list available host_info types.

  • Added host info messages for sysctl and date for WALDO.

  • Added new host info type messages and error responses.

  • Implemented fixes for various integration tests (XenServer).

  • Implemented a CentOS 7 (or systemd based Linux OS) packaging fix.

  • Implemented a fix to the non-cryptographic RNG used for mostly reconnect logic. It was not seeded, leading to agents to connect at similar times.

  • Fixed crash issue for Windows logging.

  • Implemented fix for authorization timeouts and reconnects.

  • Implemented cleanup and a bug fix for the DNS module in Luvit that caused the agent to crash.

  • Added support for CentOS 7.

  • Added support for server-side monitoring configuration.

  • Made the file related to the server-side monitoring configuration available as part of the monitoring API response.

  • Added a formal link for the Windows 64-bit installer.

  • Made latest agent version available through http://stable.packages.cloudmonitoring.rackspace.com/VERSION.

  • Implemented several minor bug fixes and reliability improvements.

Known issues#

None for this release.

v1.7, July 03, 2014#

These release notes correspond to a new-feature release of Rackspace Monitoring.

What’s new#

Server-side configuration is now available. For details, see Install and configure the agent.

Server-side configuration enables you to configure checks and alarms by using a YAML file. Users must update their monitoring agent to take advantage of this new feature.

This API-only feature is available globally without regard to region.

Resolved issues#

The following agent-specific changes were made: - Fixed a race condition in the server-side agent configuration feature

  • Added the server-side configuration directory during installation

  • Made small logging adjustments for easier reading

Known issues#

None for this release.

v1.6, June 16, 2014#

These release notes correspond to a new-feature release of Rackspace Monitoring.

What’s new#

Resolved issues#

  • Fixed Lua crash dump uploading.

  • Enabled Windows plug-in scripts to be written in Batch or PowerShell.

  • Fixed the path to agent plug-ins on Windows. The plug-in path for Windows is C:\Program Files\Rackspace Monitoring\plugins.

  • Fixed an error running plug-ins that use Windows PowerShell when there are spaces in the path.

  • Increased the check limit and the alarm limit for accounts to 1000 (increased from 200).

  • Added note to the Developer Guide that the maximum limit for consecutiveCount is five.

  • Made several minor bug fixes and reliability improvements.

Known issues#

None for this release.

v1.5, March 31, 2014#

These release notes correspond to a new-feature release of Rackspace Monitoring.

What’s new#

A new “suppressions” feature was added. This feature enables you to suspend monitoring alarms for set periods of time. See Suppressions.

Resolved issues#

Made many minor bug fixes and reliability improvements.

Documentation issues#

  • Clarified the descriptions of the alert policies to include how the polling window affects the alarm state. See Alert Policies (Consistency Level).

  • Added note that hop numbers in the response that are repeated indicate more than one hop at the same juncture (packet split).

  • Added explanation of the agent connections GUID attribute (it is a string that does not always conform to GUID formats).

Known issues#

None for this release.

v1.4, August 22, 2012#

These release notes correspond to the Unlimited Availability release of Rackspace Monitoring.

What’s new#

  • You can pass in custom headers for a remote.http check. Depending on the site that you’re testing, this can be crucial in determining if your site is up. See remote.http.

  • You can add a previous operator to the Alarm Language, which enables you to look at a metric value in the previous time period. It is useful for detecting changes. For example, to detect an SSH fingerprint change, you’d construct an Alarm snippet like the following one:

    if (metrics['fingerprint'] != previous(metrics['fingerprint'])) {
      return new AlarmStatus(CRITICAL, 'SSH fingerprint changed to #{fingerprint}');
    }
    

    For more information, see Alarm language.

  • Email alerts have been improved by including in them observations of the particular check from each monitoring zone. This enables you to get visibility on the status from any of the included monitoring zones of a check. Additionally, the message has been reformatted to look more user friendly.

    Subject:

    ** CRITICAL: host server check on web1.domain.com **
    

    Body:

    ================== Rackspace Monitoring Notification ===================
    Entity: web1.domain.com (50.56.179.42)
    Check: host server check (remote.ping)
    Alarm: Ping packet loss
    Status: CRITICAL
    Date: Mon Aug 13 2012 15:18:59 GMT+0000 (UTC)
    
    --------------------------- Observations -------------------------------------
    
    CRITICAL - Packet loss is 20%
    London (LON), 25 sec ago
    
    OK - Packet loss is normal
    London (LON), 10 sec ago
    
    OK - Packet loss is normal
    London (LON), 56 sec ago
    
    ------------------------------------------------------------------------------
    
    This is an automated Rackspace Monitoring notification. You received this email
    because you are listed as a notification recipient.
    
  • You can include the query strings uri or entityId to retrieve specific mapped Rackspace Cloud Servers. You can also retrieve a set of entities by referencing their IDs. You cannot combine retrieving both URIs and entity IDs in the same request, but you can have up to 100 of them in a single request. See also Views.

  • Friendly names were added for the alarm examples.

  • A metadata attribute was added to the alarm API.

  • You can now include no conditionals in the alarm language, which enables you to set global filters and rely on the check availability to determine the returned state.

Resolved issues#

  • Fixed unsetting optional attributes on an update (PUT).

  • Fixed the check_type to alarm mapping for delivering inconsistent and non-deterministic updates.

  • Fixed how zlib streams are handled when a gzip HTTP stream is inflated.

  • Inserted the initial OK state for the alarm change log.

  • Fixed updating the target_alias after setting the target_hostname on the initial creation, and vice versa.

Known issues#

None for this release.

v1.3, June 14, 2012#

These release notes correspond to the Limited Availability release of Rackspace Monitoring.

What’s new#

  • The new Traceroute API lets you run a traceroute from a monitoring zone to a hostname or IP address. Like all Rackspace Monitoring features, the Traceroute API is fully dual stack, supporting both IPv4 and IPv6. The Traceroute API can be used to debug networking issues between the Rackspace Monitoring collectors and your infrastructure. See the Traceroute section.

  • Metric interpolation on the alarm status string enables you to include metrics from a check in the status string returned by an alarm. For example, to include the HTTP status code from a remote.http check, your alarm could look as follows:

    if (metrics['code'] != "200") {
        return new AlarmStatus(CRITICAL, 'Bad HTTP Status: #{code}');
    }
    

    For more information, see Alarm language

  • Payload was added to the remote.http check, which enables you send a request body during a HTTP/HTTPS request. See also remote.http attributes.

  • You can specify the number of ICMP packets to send out on the remote.ping check.

  • The Certificate Authority chain used to validate SSL certificates has been updated, which will result in fewer false positives.

  • The remote.tcp check now features the body_match_metric when you specify the send_body command.

  • You can now add the remote IP address of the collector that alerted you to the webhook notification payload.

Resolved issues#

  • Fixed writable date and time fields that were supposed to be immutable.

  • Fixed the DSL string escaping to enable you to specify special characters. See the following example:

    if (metric['code'] nregex '2\\d\\d') {
      return new AlarmStatus(CRITICAL, 'Invalid status code #{code}');
    }
    

Known issues#

None for this release.

v1.2, May 10, 2012#

These release notes correspond to the post-Early Access release of Rackspace Monitoring.

What’s new#

  • You can test execute an existing check. This feature enables you to execute a check like the traditional test check, but on a check that is already running. This will not actually execute the new check in the same check context; it executes in a new context.

  • You can now get a list of alarm recipes with templating capabilities. See Alarm Examples.

  • You can now delete an entity and have the deletion cascade to all of the associated checks and alarms in the system. As a result, you can deprovision monitors quickly and easily without having to manually make the deletions.

  • You can type check alarms so that you can receive alerts if you set up an incorrect alarm, or if an alarm does not behave as expected. The alarm also validates that all the metrics you are checking are present, and if they are not present, it automatically changes the alarm state to CRITICAL.

  • More collector infrastructure was added in the IAD region.

  • A metadata field was added on the check type.

Resolved issues#

  • Fixed a defect in the networking partitioning logic that leads to more consistent operation

  • Fixed missing version parameter in next href

  • Fixed banner match to be more consistent

  • Removed the ability to use duplicate monitoring zones

Known issues#

Alarm states have a race condition. During replay it’s possible to unintentionally store duplicates.

v1.1, March 28, 2012#

These release notes correspond to the Early Access release of Rackspace Monitoring.

What’s new#

  • You can test execute a notification. This makes it easy to execute test checks and verify that you’ll receive the payload. You can also test existing notifications.

  • Support has been added for JSONP for retrieval in the API. This gives you the tools to access the API from the browser directly.

  • Rate limiting and the time window of the rate limiting are now exposed to you through headers. This helps us with transparency and lets you know when you’re getting close to reaching your account rate limits.

  • An email notification type has been added so that you can receive email on your alerts.

  • Alarm History has been renamed to Alarm Notification History.

  • A key-value hash to facilitate multiple body matches is supported.

  • Previous state has been added to the webhook payload.

  • Audits now expose the tenant ID.

  • HTTP CORS headers have been added so that you can now do cross-domain requests.

  • The notification history log and the alarm change log have been unified.

  • Support dealing with time ordered mixed results has been improved.

  • Type checking for the alarm subsystem has been improved.

  • Detection for missing metrics has been added, as well as automatically outputting a failed state when detected.

Resolved issues#

  • Fixed hashing method to prevent a collector affinity.

  • Improved the location header returned so it works in nonstandard port setups.

  • Added previous state to the webhook field.

  • Added a label to the alarm object.

  • Unified the language around the different states in the system.

Known issues#

None for this release.