Python agent release notesRSS

Notes

This release of the Python agent includes various bug fixes as described below.

The agent can be installed using easy_install/pip/distribute via the Python Package Index or can be downloaded directly from the New Relic download site.

Bug Fixes

• Handle exceptions generated by Tornado 4 AsyncHTTPClient during fetch

  Exceptions generated by a call to the Tornado 4 AsyncHTTPClient resulted in an agent instrumentation error. Exceptions generated by AsyncHTTPClient are now recorded and properly handled.

• Using gevent/eventlet on Python 3.6 resulted in a RecursionError

  Python 3.6 introduced a change to the ssl library. This change can result in a recursion error when the New Relic agent imports ssl prior to eventlet/gevent patching the ssl library (as is the case with use of the newrelic-admin script). This recursion error has been mitigated when using newrelic-admin to start applications.

Notes

This release of the Python agent includes improvements to our introductory support for Tornado 4 applications as well as bug fixes.

The agent can be installed using easy_install/pip/distribute via the Python Package Index or can be downloaded directly from the New Relic download site.

New Feature

• Tornado 4 Outbound Cross Application Tracing

  Tornado 4 Cross Application Tracing is now supported for outbound web transactions. In addition to Service Maps, Tornado 4 transaction traces will now contain links to the application it called. With this update, Cross Application Tracing is now fully supported for Tornado 4 applications.

Bug Fixes

• Django 1.10+ browser monitoring with new-style gzip middleware

  Browser monitoring was not automatically inserted when using new style Django 1.10+ gzip middleware. Browser monitoring code is now inserted as expected.

• Using motor versions 1.X generated instrumentation errors

  Motor 1.0 removed the MotorReplicaSetClient class which resulted in the New Relic Python Agent generating an instrumentation error when attempting to access that class. The instrumentation has been fixed.

Notes

This release of the Python agent includes bug fixes and improvements to the precision of the duration data recorded for transaction traces.

The agent can be installed using easy_install/pip/distribute via the Python Package Index or can be downloaded directly from the New Relic download site.

Improved Feature

• Transaction trace precision

  This release improves the precision of transaction traces. Duration of functions in a transaction trace are now recorded at sub-millisecond precision.

Bug Fixes

• Custom classmethod traces may incorrectly return the parent class

  Wrapping classmethods in classes that are subclassed could cause the subclassed classmethod to be improperly bound to the parent class. This would result in the classmethod returning an instance of the parent class when the method was called on a child class.

• Handle incomplete Tornado coroutine traces

  Tracing a Tornado coroutine that never completed and was later garbage collected, could result in an application hang. This has been fixed.

Notes

This release of the Python agent includes a bug fix for our introductory support for Tornado 4 applications.

The agent can be installed using easy_install/pip/distribute via the Python Package Index or can be downloaded directly from the New Relic download site.

Bug Fixes

• Ending Tornado 4 transactions inside of a traced function results in an internal agent error

  Ending a Tornado 4 web transaction within a traced function previously resulted in the generation of an agent traceback. The agent now completes tracing the function before ending the transaction, resulting in the correct reporting of these transactions to New Relic.

Notes

This release of the Python agent improves our introductory support for Tornado 4 applications by adding Cross Application Tracing for web transactions that are sent to a Tornado 4 application.

The agent can be installed using easy_install/pip/distribute via the Python Package Index or can be downloaded directly from the New Relic download site.

New Feature

• Tornado 4 applications will now appear in Service Maps

  Tornado 4 Cross Application Tracing is now supported for inbound web transactions. In addition to Service Maps, the calling application's APM will now have links to the Tornado 4 application it called in the External Services tab and in Transaction Traces.

Bug Fixes

• Record exceptions outside of view handlers in Django 1.10

  Unhandled exceptions that occurred outside of a view handler in Django 1.10 were not reported. In addition to middleware exceptions, these unhandled exceptions can be generated when a view handler returns a TemplateResponse and rendering that template would result in an exception.

• Consistent WSGI/Application and WSGI/Response segments

  Prior to this release, the agent would group segments under WSGI/Application and WSGI/Response differently, depending on whether or not browser monitoring was enabled. This has now been made consistent.

• Improved New Relic WSGI middleware behavior

  The New Relic agent is now resilient when called by non-conforming middleware. Transactions will continue to be reported normally when the New Relic agent is called by WSGI middleware that does not call the close() method as described in the PEP 333 specification.

Notes

This release of the Python agent is a hotfix release to address a problem with the Tornado 4 instrumentation for curl_httpclient introduced in our last release (v2.80.0.60). Under the circumstances described below, calling CurlAsyncHTTPClient.fetch_impl() could result in an agent error.

The agent can be installed using easy_install/pip/distribute via the Python Package Index or can be downloaded directly from our download site.

Bug Fix

• Fix CurlAsyncHTTPClient.fetch_impl() instrumentation

  If fetch_impl() was called by passing the request argument positionally, but passing callback as a keyword argument, then the agent would throw an error. In practice, this only affected customers who implemented a custom version of fetch_impl(), but for those affected customers, this is now fixed.

Notes

This release of the Python agent enhances our introductory support for Tornado 4, most notably by adding support for tornado.curl_httpclient.

The agent can be installed using easy_install/pip/distribute via the Python Package Index or can be downloaded directly from the New Relic download site.

New Features

• Tornado 4 curl_httpclient support

  Previously, the agent's Tornado 4 instrumentation only supported the default tornado.simple_httpclient. With this release, support has been added for tornado.curl_httpclient as well.

• Tornado 4 ZeroMQ IOLoop support

  Using ZMQIOLoop in place of the default Tornado IOLoop is now supported for Tornado 4 applications.

Improved Feature

• Improved transaction names for Django REST Framework

  When using the Python agent with Django REST Framework and Python 3, function-based views wrapped with the rest_framework.decorators.api_view decorator would result in transactions with names based on the decorator (WrappedAPIView.handler) rather than names based on the original view function. (This did not affect applications running with Python 2.)

  With this release, the transaction names for these wrapped views are now based on the name of the original view function in both Python 2 and Python 3.

Bug Fix

• End transactions correctly with Tornado 4 @asynchronous decorator

  When the agent records a transaction for a Tornado 4 RequestHandler method that is decorated with the @tornado.web.asynchronous decorator, the transaction should remain open until RequestHandler.finish() is called. Prior to this release, there were some circumstances where the agent closed the transaction prematurely, resulting in transaction durations that were too short. This has now been fixed.

Notes

This release of the Python agent adds support for reporting datastore instance information for the elasticsearch library.

Over the past few releases, support has been added to the agent for collecting datastore instance information for a number of drivers. See a complete list of supported packages.

The agent can be installed using easy_install/pip/distribute via the Python Package Index or can be downloaded directly from the New Relic download site.

New Feature

• Datastore instance information for elasticsearch

  For operations made using the official Elasticsearch client, the Python agent will collect the host and port of the elasticsearch server. This data will be displayed in transaction traces and in slow query traces in APM.

Bug Fix

• Cornice 2.3.0 TypeError

  When using the latest version of Cornice,the Python agent would raise a TypeError exception due to not handling the changed signature of cornice.Service.decorate_view(). Earlier versions of Cornice were not affected, and with this release, the Python agent no longer raises an error for version 2.3.0.

Notes

This release of the Python agent adds support for reporting datastore instance information for the python-memcached library. It also addresses a couple of Python 3 issues related to naming metrics and transactions.

Over the past few releases, support has been added to the agent for collecting datastore instance information for a number of drivers. See a complete list of supported packages.

The agent can be installed using easy_install/pip/distribute via the Python Package Index or can be downloaded directly from the New Relic download site.

New Features

• Datastore instance information for python-memcached

  For most memcached operations made using the python-memcached package, the Python agent will collect the host and port of the memcached server, as well as the memcached database index number. This data will be displayed in transaction traces and in slow query traces in APM.

  Supported operations include: add, append, cas, decr, delete, get, gets, incr, prepend, replace, set.

  Currently, the Python agent does not collect instance information for operations that interact with multiple memcached instances. These include: delete_multi, flush_all, get_multi, get_slabs, get_stats, set_multi.

Bug Fixes

• Names for subclass methods in Python 3

  Previously, when using Python 3, if a method was defined on a base class, but was called as a method on an instance of a subclass, the agent would generate a metric name or transaction name for that method using the base class. It would also use the module name of the base class in the metric or transaction name.

  With this release, the metric name or transaction name will use the class name of the subclass, as well as the module name of the subclass. This is consistent with how the naming works for Python 2.

• Names for Flask pluggable views in Python 3

  When monitoring Python 3 Flask applications, the agent would create metric and transaction names for Flask pluggable (or, class-based) views using the name of the View.as_view() classmethod, making it impossible to distinguish between different views in APM. Now, the metric and transaction name uses the name that is passed into View.as_view().

  For example, a view defined as view_func=ShowUsersView.as_view('show_users') would use show_users as the view name. This is consistent with how the naming works for Python 2.

• Redis client with custom connection pool

  If a custom connection pool was used with the redis client, the agent could throw an AttributeError when attempting to gather datastore instance information. One library known to define a custom connection pool that encountered this issue is Redis Blaster. With this release, the agent will no longer throw an error for custom connection pools. In addition, the agent will now be able to gather datastore instance information for some operations, even when using a custom connection pool.

• Developer Mode error message during "fake" harvest

  When using the configuration setting of developer_mode = True, the agent does not transmit any data to New Relic. Instead, it performs a "fake harvest". In some cases, it was possible for the agent to log an AssertionError during the "fake harvest", but this issue has now been fixed.