Merge branch 'main' into feat/golive-form-updates

.devcontainer/Dockerfile

@@ -17,6 +17,7 @@ RUN apt-get update \
     emacs \
     exa \
     fd-find \
+    fzf \
     git \
     iputils-ping \
     iproute2 \

.devcontainer/scripts/installations.sh

@@ -13,6 +13,10 @@ echo -e "alias ll='exa -alh@ --git'" >> ~/.zshrc
 echo -e "alias lt='exa -al -T -L 2'" >> ~/.zshrc
 echo -e "alias poe='poetry run poe'" >> ~/.zshrc
 
+echo -e "# fzf key bindings and completion" >> ~/.zshrc
+echo -e "source /usr/share/doc/fzf/examples/key-bindings.zsh" >> ~/.zshrc
+echo -e "source /usr/share/doc/fzf/examples/completion.zsh" >> ~/.zshrc
+
 # Poetry autocomplete
 echo -e "fpath+=/.zfunc" >> ~/.zshrc
 echo -e "autoload -Uz compinit && compinit"

gunicorn_config.py

@@ -1,11 +1,13 @@
 import os
 import sys
+import time
 import traceback
 
 import gunicorn  # type: ignore
 import newrelic.agent  # See https://bit.ly/2xBVKBH
 
-newrelic.agent.initialize()  # noqa: E402
+environment = os.environ.get("NOTIFY_ENVIRONMENT")
+newrelic.agent.initialize(environment=environment)  # noqa: E402
 
 # Guincorn sets the server type on our app. We don't want to show it in the header in the response.
 gunicorn.SERVER = "Undisclosed"
@@ -20,10 +22,34 @@
 # to be larger than the idle timeout configured for the load balancer.
 # > By default, Elastic Load Balancing sets the idle timeout value for your load balancer to 60 seconds.
 # https://docs.aws.amazon.com/elasticloadbalancing/latest/application/application-load-balancers.html#connection-idle-timeout
-on_aws = os.environ.get("NOTIFY_ENVIRONMENT", "") in ["production", "staging", "scratch", "dev"]
+on_aws = environment in ["production", "staging", "scratch", "dev"]
 if on_aws:
     keepalive = 75
 
+    # The default graceful timeout period for Kubernetes is 30 seconds, so
+    # make sure that the timeouts defined here are less than the configured
+    # Kubernetes timeout. This ensures that the gunicorn worker will exit
+    # before the Kubernetes pod is terminated. This is important because
+    # Kubernetes will send a SIGKILL to the pod if it does not terminate
+    # within the grace period. If the worker is still processing requests
+    # when it receives the SIGKILL, it will be terminated abruptly and
+    # will not be able to finish processing the request. This can lead to
+    # 502 errors being returned to the client.
+    #
+    # Also, some libraries such as NewRelic might need some time to finish
+    # initialization before the worker can start processing requests. The
+    # timeout values should consider these factors.
+    #
+    # Gunicorn config:
+    # https://docs.gunicorn.org/en/stable/settings.html#graceful-timeout
+    #
+    # Kubernetes config:
+    # https://kubernetes.io/docs/concepts/containers/container-lifecycle-hooks/
+    graceful_timeout = 85
+    timeout = 90
+
+# Start timer for total running time
+start_time = time.time()
 
 def on_starting(server):
     server.log.info("Starting Notifications Admin")
@@ -36,7 +62,9 @@ def worker_abort(worker):
 
 
 def on_exit(server):
+    elapsed_time = time.time() - start_time
     server.log.info("Stopping Notifications Admin")
+    server.log.info("Total gunicorn Admin running time: {:.2f} seconds".format(elapsed_time))
 
 
 def worker_int(worker):

newrelic.ini

@@ -180,6 +180,16 @@ browser_monitoring.auto_instrument = false
 # call tree.
 # thread_profiler.enabled = true
 
+# If this setting is enabled, it will capture package and version
+# information on startup of the agent that is displayed in the APM
+# environment tab.
+# In applications that have a large number of packages, having this
+# setting enabled may cause a CPU spike as it captures all the package
+# and version information. It is recommended in those cases to disable
+# this setting.
+# Disabling this setting will disable the ability to detect vulnerabilities in outdated packages.
+package_reporting.enabled = false
+
 # ---------------------------------------------------------------------------
 
 #
@@ -188,7 +198,7 @@ browser_monitoring.auto_instrument = false
 # specific environment will be used when the environment argument to the
 # newrelic.agent.initialize() function has been defined to be either
 # "development", "test", "staging" or "production".
-#
+
 
 [newrelic:development]
 # monitor_mode = false

pyproject.toml

@@ -58,7 +58,7 @@ redis = "3.5.3" # pinned for now
 xlrd = "1.2.0" # this is pinned so that we can continue to support .xslm files; see https://github.com/pyexcel/pyexcel-xls/issues/53
 # Putting upgrade on hold due to v1.0.0 using sha512 instead of sha1 by default
 itsdangerous = "2.2.0"  # pyup: <1.0.0
-newrelic = "8.10.0" # Pinned to 8.10.0 for now, 8.11.0 caused a performance regression: https://gcdigital.slack.com/archives/C012W5K734Y/p1709668046344929
+newrelic = "10.3.0"
 aws-xray-sdk = "^2.14.0"
 
 validators = "^0.28.3"