feat: Change the logging config for edx-platform.

[dianakhuang]

The logs changed between ec2 and k8s because the config between the two is different. This tries to re-introduce some of the extra logging parameters and adds in the DD trace id for good measure.

https://2u-internal.atlassian.net/browse/BOMS-661

[Copilot]

Pull request overview

This PR adjusts the generated docker-production.py Django settings inside the edx-platform image to bring k8s logging behavior closer to the prior EC2 configuration, including adding Datadog trace correlation data to console logs.

Changes:

• Adds an extended console log formatter with hostname/process/user/IP/file/line context.
• Includes Datadog dd.trace_id in the console log format.
• Leaves the tracking log handler behavior intact while adding formatting overrides.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated no new comments.

Comments suppressed due to low confidence (1)

dockerfiles/edx-platform.Dockerfile:292

• The new console formatter references custom LogRecord keys (e.g. dd.trace_id, userid, remoteip). If any log record doesn’t provide these fields, Python logging will raise a formatting error and drop the log message. To make this robust for third-party / early-startup logs, add defaults to the formatter config so missing keys have safe fallbacks.

syslog_format = ("[%(name)s] %(levelname)s [trace_id %(dd.trace_id)s]"
                     "[{hostname}  %(process)d] [user %(userid)s] [ip %(remoteip)s] [%(filename)s:%(lineno)d] "
                     "- %(message)s").format(hostname=platform.node().split(".")[0])

LOGGING["formatters"]["syslog_format"] = {"format": syslog_format}

[Copilot]

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated no new comments.

Comments suppressed due to low confidence (1)

dockerfiles/edx-platform.Dockerfile:292

• The console formatter includes %(dd.trace_id)s whenever ddtrace is importable, but log records won’t necessarily contain dd.trace_id unless ddtrace log injection is enabled (and even then it may be absent in some contexts). Python’s logging formatter will raise a KeyError for missing format fields, which can break logging at runtime. Consider (a) gating inclusion of the trace id on ddtrace.config.logs_injection and (b) providing defaults for non-standard fields (including dd.trace_id, userid, and remoteip) to prevent formatter errors. This also makes the ddtrace/non-ddtrace formats consistent for the hostname/process portion.

try:
    # This should check to see if ddtrace is available in the context for setting up the loggers.
    # If it is available, include the trace_id in the log string. If it is not, exclude it.
    import ddtrace
    syslog_format = ("[%(name)s] %(levelname)s [trace_id %(dd.trace_id)s] "

[Copilot]

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 1 comment.

[Copilot]

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated 1 comment.

Comments suppressed due to low confidence (1)

dockerfiles/edx-platform.Dockerfile:292

• The presence of the ddtrace package alone doesn’t guarantee that dd.* fields are injected into every LogRecord. If ddtrace is installed but logs injection is disabled, this formatter can raise KeyError at runtime when it tries to render %(dd.trace_id)s, etc. Consider guarding on ddtrace’s logs-injection setting (or an explicit env flag) and falling back to the non-DD format when injection isn’t enabled.

try:
    # This should check to see if ddtrace is available in the context for setting up the loggers.
    # If it is available, include Datadog information in the log string. If it is not, exclude it.
    import ddtrace
    syslog_format = ("[%(name)s] %(levelname)s "

[Copilot]

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated no new comments.

Comments suppressed due to low confidence (1)

dockerfiles/edx-platform.Dockerfile:296

• The new syslog_format references non-standard LogRecord keys (e.g., userid, remoteip, and the dd.* fields). With the default logging.Formatter, any log record missing these keys will raise a formatting KeyError and can cause log lines to be dropped; additionally, importing ddtrace just to check availability can add import-time overhead/side effects.

Consider (1) checking for ddtrace without importing it, and (2) configuring formatter defaults so missing fields don’t break logging.

    syslog_format = ("[%(name)s] %(levelname)s "
                    "[dd.service=%(dd.service)s dd.env=%(dd.env)s dd.version=%(dd.version)s dd.trace_id=%(dd.trace_id)s"
                     " dd.span_id=%(dd.span_id)s] "
                     "[{hostname}] [process %(process)d] [user %(userid)s] [ip %(remoteip)s] [%(filename)s:%(lineno)d] "
                     "- %(message)s").format(hostname=platform.node().split(".")[0])

[Copilot]

Pull request overview

Copilot reviewed 1 out of 1 changed files in this pull request and generated no new comments.

Comments suppressed due to low confidence (1)

dockerfiles/edx-platform.Dockerfile:292

• The new syslog formatter references non-standard LogRecord fields (e.g., userid, remoteip, and in the ddtrace branch dd.trace_id / dd.span_id). Python logging raises KeyError during formatting if any referenced field is missing, which can cause log lines to be dropped and emit noisy "Logging error" traces. Consider adding a small logging Filter to supply safe defaults for these fields (and avoid importing ddtrace just to check availability).

try:
    # This should check to see if ddtrace is available in the context for setting up the loggers.
    # If it is available, include Datadog information in the log string. If it is not, exclude it.
    import ddtrace
    syslog_format = ("[%(name)s] %(levelname)s "