#AE_FIELDS reference

Verified field map for the agent-events journal namespace. Every claim here is traceable to producer source at <repo>/src/daemon/status-file.c (the schema is STATUS_FILE_VERSION = 28, src/daemon/status-file.h:14).

The .local draft agent-events-journals.md was found to have 14 high-severity divergences (wrong enums, missing fields, misattributed semantics). This file supersedes it.

#How journal field names are formed

Producer JSON path -> journal field name:

1. Producer emits a JSON path (e.g. agent.profile[0]).
2. log2journal applies --prefix 'AE_' (literal prepend; the prefix is NOT transliterated -- src/collectors/log2journal/log2journal-help.c:108-109).
3. log2journal walks the JSON tree. For nested objects, it joins parent + child with _. For arrays, it appends _<index> (log2journal-json.c:477-511).
4. Per-character transliteration applies a 256-entry map (log2journal.c:8-61): lowercase -> uppercase, digits and uppercase pass through, everything else (including ., -, :, /, @, _, [, ]) maps to _. Consecutive underscores are collapsed (log2journal-json.c:395-396).

So: - agent.id -> AE_AGENT_ID - agent.profile[0] -> AE_AGENT_PROFILE_0 - host.boot.id -> AE_HOST_BOOT_ID - @timestamp -> AE__TIMESTAMP (note: double underscore because @ -> _) - fatal.errno -> AE_FATAL_ERRNO (NOT from a top-level fatal_errno; the path is nested)

#Top-level POST-time fields (always present)

Added by post_status_file() at status-file.c:967-976. These sit at the JSON top level, so they have NO nested-path prefix in the journal field name.

JSON path	Journal field	Type	Always?	Triage meaning
exit_cause	AE_EXIT_CAUSE	string	yes	Human-readable label for why this session ended. The first thing to look at. See enum below.
message	AE_MESSAGE	string	yes	One-line summary; subject of FTS search.
priority	AE_PRIORITY	int	yes	Syslog priority (lower = more severe).
version_saved	AE_VERSION_SAVED	uint	yes	The status file's own version (typically 28).
agent_version_now	AE_AGENT_VERSION_NOW	string	yes	Posting agent's version (the agent that did the POST = next session, NOT the one that crashed).
agent_pid_now	AE_AGENT_PID_NOW	uint	yes	PID of the agent that POSTed.
host_memory_critical	AE_HOST_MEMORY_CRITICAL	bool	yes	Was the host under memory pressure at POST time?
host_memory_free_percent	AE_HOST_MEMORY_FREE_PERCENT	uint	yes	% free RAM at POST time.
agent_health	AE_AGENT_HEALTH	string	yes	Health classification across restart history. See enum below.
@timestamp	AE__TIMESTAMP	RFC3339	yes	When the captured session ended.
version	AE_VERSION	uint	yes	The schema version (STATUS_FILE_VERSION). Use to slice by schema.

These fields are always present on every record and are the safest scoping anchors. AE_EXIT_CAUSE, AE_AGENT_HEALTH, AE_VERSION are all good index-friendly facets.

#agent.* fields (dsf_json_agent, status-file.c:99-156)

The previous (crashed) session's agent state.

JSON path	Journal field	Type	Version-gating	Triage meaning
agent.id	AE_AGENT_ID	UUID	always	Netdata machine GUID (per-install, persistent). The "agent identity". DIFFERENT from host.id.
agent.since	AE_AGENT_SINCE	RFC3339	v>=24	When this install was first registered.
agent.ephemeral_id	AE_AGENT_EPHEMERAL_ID	UUID	always	Unique per Netdata invocation (changes on every restart). Use to group multiple events from the same crashed session.
agent.version	AE_AGENT_VERSION	string	always	The version of the crashed session. Slice on this for regression-spotting.
agent.uptime	AE_AGENT_UPTIME	int seconds	always	Duration the crashed session ran. Short uptime + crash = startup bug.
agent.node_id	AE_AGENT_NODE_ID	UUID	always	Cloud node UUID (empty when agent isn't claimed).
agent.claim_id	AE_AGENT_CLAIM_ID	UUID	always	Cloud claim UUID (empty when not claimed). Presence -> cloud-connected agent.
agent.restarts	AE_AGENT_RESTARTS	uint	always	Total restart count for this install. High value + recent crash = agent loop.
agent.crashes	AE_AGENT_CRASHES	uint	v>=24	Total crash count.
agent.pid	AE_AGENT_PID	uint	v>=27	PID of the crashed session.
agent.posts	AE_AGENT_POSTS	uint	v>=22	Total POSTs from this install.
agent.aclk	AE_AGENT_ACLK	enum	v>=22	Cloud connection state. See enum.
agent.profile[N]	AE_AGENT_PROFILE_0..N	enum array	always	Bitmap rendered as array. Slice by parent vs child vs iot. See enum.
agent.status	AE_AGENT_STATUS	enum	always	DAEMON_STATUS at the time of capture. See enum.
agent.exit_reason[N]	AE_AGENT_EXIT_REASON_0..N	enum array	always	EXIT_REASON bitmap rendered as array. Empty array = no specific reason. See enum.
agent.install_type	AE_AGENT_INSTALL_TYPE	string	always	kickstart, binpkg, static, etc. Slice on this for "is this a packaging issue?"
agent.db_mode	AE_AGENT_DB_MODE	string	v>=14	dbengine memory mode.
agent.db_tiers	AE_AGENT_DB_TIERS	uint	v>=14	Number of dbengine tiers.
agent.kubernetes	AE_AGENT_KUBERNETES	bool	v>=14	Kubernetes deployment? Slice on this for k8s-specific issues.
agent.sentry_available	AE_AGENT_SENTRY_AVAILABLE	bool	v>=16	Is Sentry enabled?
agent.reliability	AE_AGENT_RELIABILITY	int	always	Signed reliability counter (positive = healthy run streak; negative = crash streak). <= -2 -> crash-loop.
agent.stack_traces	AE_AGENT_STACK_TRACES	string	always	Backtrace backend name (libbacktrace, none).
agent.timings.init	AE_AGENT_TIMINGS_INIT	int seconds	always	How long startup took. Long init + crash = startup bug.
agent.timings.exit	AE_AGENT_TIMINGS_EXIT	int seconds	always	How long shutdown took.

#metrics.* fields (dsf_json_metrics, :158-193)

Snapshot of the database at the time of capture. Useful for "big-database crashes" investigations.

JSON path	Journal field
metrics.nodes.total	AE_METRICS_NODES_TOTAL
metrics.nodes.receiving	AE_METRICS_NODES_RECEIVING
metrics.nodes.sending	AE_METRICS_NODES_SENDING
metrics.nodes.archived	AE_METRICS_NODES_ARCHIVED
metrics.metrics.collected	AE_METRICS_METRICS_COLLECTED
metrics.metrics.available	AE_METRICS_METRICS_AVAILABLE
metrics.instances.collected	AE_METRICS_INSTANCES_COLLECTED
metrics.instances.available	AE_METRICS_INSTANCES_AVAILABLE
metrics.contexts.collected	AE_METRICS_CONTEXTS_COLLECTED
metrics.contexts.available	AE_METRICS_CONTEXTS_AVAILABLE

#host.* fields (dsf_json_host, :195-253)

Host-level info, mostly stable across crashes on the same host.

JSON path	Journal field	Triage meaning
host.id	AE_HOST_ID	OS-level /etc/machine-id. Different from AE_AGENT_ID (Netdata's own identifier). Use to spot multiple agents on the same host.
host.architecture	AE_HOST_ARCHITECTURE	x86_64, aarch64, armv7l, ... Slice for arch-specific bugs.
host.virtualization	AE_HOST_VIRTUALIZATION	none, kvm, vmware, lxc, docker, ...
host.container	AE_HOST_CONTAINER	none, docker, kubernetes, ...
host.uptime	AE_HOST_UPTIME	MISLEADING NAME. Stores boottime EPOCH (status-file.c:202 writes ds->boottime from now_boottime_sec()). NOT a duration. Compute uptime via now - AE_HOST_UPTIME.
host.timezone	AE_HOST_TIMEZONE	string, v>=20
host.cloud_provider	AE_HOST_CLOUD_PROVIDER	aws, gcp, azure, ..., v>=20
host.cloud_instance	AE_HOST_CLOUD_INSTANCE	EC2 instance type etc., v>=20
host.cloud_region	AE_HOST_CLOUD_REGION	v>=20
host.system_cpus	AE_HOST_SYSTEM_CPUS	uint. Slice for "low-cpu environment" bugs.
host.boot.id	AE_HOST_BOOT_ID	UUID, changes on every host boot.
host.memory.total	AE_HOST_MEMORY_TOTAL	bytes, only when OS_SYSTEM_MEMORY_OK
host.memory.free	AE_HOST_MEMORY_FREE	bytes
host.memory.netdata	AE_HOST_MEMORY_NETDATA	bytes used by netdata, v>=21
host.memory.oom_protection	AE_HOST_MEMORY_OOM_PROTECTION	uint, v>=21
host.disk.db.total	AE_HOST_DISK_DB_TOTAL	bytes available to dbengine
host.disk.db.free	AE_HOST_DISK_DB_FREE	bytes free
host.disk.db.inodes_total	AE_HOST_DISK_DB_INODES_TOTAL	uint
host.disk.db.inodes_free	AE_HOST_DISK_DB_INODES_FREE	uint
host.disk.db.read_only	AE_HOST_DISK_DB_READ_ONLY	bool. True + crash -> "disk read-only" cause.
host.disk.netdata.dbengine	AE_HOST_DISK_NETDATA_DBENGINE	bytes used by dbengine files.
host.disk.netdata.sqlite	AE_HOST_DISK_NETDATA_SQLITE	bytes used by SQLite files.
host.disk.netdata.other	AE_HOST_DISK_NETDATA_OTHER	bytes used by other files.
host.disk.netdata.last_updated	AE_HOST_DISK_NETDATA_LAST_UPDATED	RFC3339.

#os.* fields (dsf_json_os, :255-266)

JSON path	Journal field	Triage meaning
os.type	AE_OS_TYPE	enum: unknown, linux, freebsd, macos, windows.
os.kernel	AE_OS_KERNEL	Kernel version string.
os.name	AE_OS_NAME	Distro name (e.g. Ubuntu, CentOS Stream).
os.version	AE_OS_VERSION	Distro version.
os.family	AE_OS_FAMILY	os_id (e.g. ubuntu). Slice for distro-specific issues.
os.platform	AE_OS_PLATFORM	os_id_like (parent distro family, e.g. debian). NOT a rewrite of AE_OS_FAMILY -- they're independent producer fields.

#hw.* fields (dsf_json_hw, :268-319)

DMI / SMBIOS data. Useful for hardware-specific bug investigation. Privacy-sensitive serials and asset_tags are commented out at the producer side (status-file.c:275-276, :294-295, :304-305) and never reach the journal.

JSON path	Journal field	Notes
hw.sys.vendor	AE_HW_SYS_VENDOR	BIOS / system vendor.
hw.sys.uuid	AE_HW_SYS_UUID	System UUID.
hw.product.name	AE_HW_PRODUCT_NAME	Product name (e.g. MacBookPro18,3).
hw.product.version	AE_HW_PRODUCT_VERSION
hw.product.sku	AE_HW_PRODUCT_SKU
hw.product.family	AE_HW_PRODUCT_FAMILY
hw.board.name	AE_HW_BOARD_NAME
hw.board.version	AE_HW_BOARD_VERSION
hw.board.vendor	AE_HW_BOARD_VENDOR
hw.chassis.type	AE_HW_CHASSIS_TYPE	Numeric (e.g. 6 = desktop, 9 = laptop).
hw.chassis.vendor	AE_HW_CHASSIS_VENDOR
hw.chassis.version	AE_HW_CHASSIS_VERSION
hw.bios.date	AE_HW_BIOS_DATE
hw.bios.release	AE_HW_BIOS_RELEASE
hw.bios.version	AE_HW_BIOS_VERSION
hw.bios.vendor	AE_HW_BIOS_VENDOR

#product.* fields (dsf_json_product, :321-329)

JSON path	Journal field
product.vendor	AE_PRODUCT_VENDOR
product.name	AE_PRODUCT_NAME
product.type	AE_PRODUCT_TYPE

#fatal.* fields (dsf_json_fatal, :331-367)

Present on crashes and deliberate fatal conditions. Empty on graceful exits.

JSON path	Journal field	Version-gating	Triage meaning
fatal.line	AE_FATAL_LINE	always	Source line of the panic. Combine with FILENAME and FUNCTION for de-dup.
fatal.filename	AE_FATAL_FILENAME	always	Source file. Slice on this for "this file is buggy".
fatal.function	AE_FATAL_FUNCTION	always	Function name (with demangled symbol). Slice on this for "this function is buggy".
fatal.message	AE_FATAL_MESSAGE	always	Panic message. Subject of FTS.
fatal.errno	AE_FATAL_ERRNO	always	errno string at panic.
fatal.thread	AE_FATAL_THREAD	always	Worker thread name (e.g. CTXLOAD, STREAM:63).
fatal.thread_id	AE_FATAL_THREAD_ID	always	POSIX TID.
fatal.stack_trace	AE_FATAL_STACK_TRACE	always	Backtrace. Real addresses preserved (anonymization is dedup-only, status-file-dedup.c:26-36).
fatal.signal_code	AE_FATAL_SIGNAL_CODE	v>=16	SIGNAL/SI_CODE formatted (e.g. SIGSEGV/SEGV_MAPERR). Empty -> not a signal crash. Primary signal-crash predicate. See enum.
fatal.sentry	AE_FATAL_SENTRY	v>=17	Was a Sentry submission attempted?
fatal.fault_address	AE_FATAL_FAULT_ADDRESS	v>=18	Hex address of the fault. Empty when signal_code == 0.
fatal.worker_job_id	AE_FATAL_WORKER_JOB_ID	v>=23	Worker job ID at panic.

#Enum reference

#AE_AGENT_STATUS (DAEMON_STATUS)

Source: src/daemon/status-file.c:23-33.

Value	Meaning for triage
none	No prior status (very first session).
initializing	Crashed during startup -> startup bug. Combine with agent.timings.init for context.
running	Crashed during normal operation -> the most "interesting" class.
exiting	Crashed during shutdown -> shutdown-path bug.
exited	Graceful exit (no crash).

#AE_AGENT_ACLK (CLOUD_STATUS)

Source: src/claim/cloud-status.c:5-15.

Value	Meaning for triage
available	Default; not yet attempted.
online	Connected to Cloud (ACLK up).
indirect	Connected via parent.
banned	Cloud rejected (claim issue).
offline	Disconnected (network or shutdown).

(The .local draft listed a disabled value -- it does NOT exist in the producer source.)

#AE_AGENT_HEALTH

Source: src/daemon/status-file.c:929-952. Computed by the agent (not the ingestion server) at POST time across restart history. Used to isolate crash classes.

Value	Meaning for triage
healthy-first	First run, no prior crashes. Boring (filter out).
healthy-repeated	Multiple healthy runs in a row.
healthy-loop	Reliability >= 2 consecutive healthy runs.
healthy-recovered	Was unhealthy, now healthy.
crash-first	First crash ever on this install. Interesting -- new bug?
crash-entered	Single crash, then recovered.
crash-loop	Reliability <= -2 (repeated crashes). Highest-priority class.
crash-repeated	Two or more crashes.

To find ALL crashes: (AE_AGENT_HEALTH in crash-first, crash-loop, crash-repeated, crash-entered).

#AE_AGENT_PROFILE_* (ND_PROFILE bitmap)

Source: src/daemon/config/netdata-conf-profile.c:7-15.

Value	Meaning
standalone	Single-node deployment.
parent	Streaming parent. Slice for "parent-only" bugs.
child	Streaming child. Slice for "child-only" bugs.
iot	IoT / lightweight profile.

(The .local draft listed dopple and store-child -- they do NOT exist; iot was missing.)

#AE_AGENT_EXIT_REASON_* (EXIT_REASON bitmap)

Source: src/libnetdata/exit/exit_initiated.c:7-38. The EXIT_REASON bitmap renders as a JSON array. Empty bitmap -> empty array (no none element).

20 distinct strings:

Value	Meaning
signal-segmentation-fault	SIGSEGV received.
signal-bus-error	SIGBUS received.
signal-floating-point-exception	SIGFPE received.
signal-illegal-instruction	SIGILL received.
signal-abort	SIGABRT received (assertion / abort()).
signal-bad-system-call	SIGSYS received.
signal-cpu-time-limit-exceeded	SIGXCPU received.
signal-file-size-limit-exceeded	SIGXFSZ received.
signal-quit	SIGQUIT received.
signal-terminate	SIGTERM received (graceful kill).
signal-interrupt	SIGINT received (Ctrl-C).
out-of-memory	OOM panic.
already-running	Another instance held the listen socket.
fatal	Generic fatal() call.
api-quit	API endpoint requested exit.
cmd-exit	Explicit netdata --exit invocation.
service-stop	Service manager (systemd) sent stop.
system-shutdown	Host shutting down.
update	Replaced by a new version.
shutdown-timeout	Shutdown took too long.

(The .local draft was significantly wrong here -- listed ~10 invented values like exit-called, exit-and-update, cannot-allocate, oom, assertion-failed, none of which exist in producer source.)

#AE_EXIT_CAUSE (top-level)

Source: src/daemon/status-file.c:1097-1286. Computed by the agent, NOT the ingestion server. The most useful field for classifying records.

26 distinct strings:

Initial / no prior state (1):

Value	Meaning
no last status	First-ever start; no prior status file readable.

Prior was EXITED (graceful) (7):

Value	Meaning
exit no reason	Prior exited cleanly with no reason recorded.
deadly signal and exit	Got a deadly signal but exited normally.
fatal and exit	Hit a fatal but managed to exit.
exit on system shutdown	Host shutting down; agent stopped gracefully.
exit to update	Stopped to allow an update.
exit and updated	Stopped and was replaced by a new version.
exit instructed	netdata --exit or service stop.

Prior was INITIALIZING (8):

Value	Meaning
abnormal power off	Power loss during startup.
deadly signal on start	Signal during startup.
out of memory	OOM during startup. (.local draft says cannot allocate -- wrong.)
already running	Listen socket conflict at init.
disk read-only	Filesystem read-only at init.
disk full	Disk full at init.
disk almost full	Disk near capacity at init.
fatal on start	fatal() during startup.
killed hard on start	SIGKILL/SIGTERM during startup.

Prior was EXITING (5):

Value	Meaning
deadly signal on exit	Signal during shutdown.
exit timeout	Shutdown didn't complete in time.
fatal on exit	fatal() during shutdown.
killed hard on shutdown	SIGKILL during shutdown (host shutdown).
killed hard on update	SIGKILL during hot update.
killed hard on exit	SIGKILL during exit.

Prior was RUNNING (6):

Value	Meaning
abnormal power off	Power loss during normal operation.
out of memory	OOM during normal operation.
deadly signal	Signal received during normal operation.
killed fatal	SIGKILL after a fatal.
killed hard low ram	OOM-killed (RAM pressure).
killed hard	SIGKILL/SIGTERM from outside.

#AE_OS_TYPE (DAEMON_OS_TYPE)

Source: src/daemon/status-file.c:35-45.

unknown, linux, freebsd, macos, windows.

#AE_FATAL_SIGNAL_CODE

Format: SIGNAL/SI_CODE (e.g. SIGSEGV/SEGV_MAPERR). Sources: src/libnetdata/signals/signal-code.c:12-53 (signal name map), :97-184 (per-signal SI_CODE map).

Most relevant for crash triage:

Value	Meaning
SIGSEGV/SEGV_MAPERR	Invalid memory map (NULL pointer, freed memory).
SIGSEGV/SEGV_ACCERR	Access violation (write to read-only page).
SIGSEGV/SEGV_BNDERR	Address bound check fault.
SIGSEGV/SEGV_PKUERR	Protection key fault.
SIGBUS/BUS_ADRALN	Alignment error.
SIGBUS/BUS_ADRERR	Non-existent physical address.
SIGBUS/BUS_OBJERR	Object-specific bus error.
SIGFPE/FPE_INTDIV	Integer divide by zero.
SIGFPE/FPE_INTOVF	Integer overflow.
SIGFPE/FPE_FLTDIV	Float divide by zero.
SIGABRT/SI_TKILL	abort() / assertion failure (typical SI_CODE for abort()).
SIGTRAP/TRAP_BRKPT	Breakpoint trap.
SIGTRAP/TRAP_TRACE	Trace trap.

Empty AE_FATAL_SIGNAL_CODE -> not a signal crash (it's a deliberate fatal or a graceful exit).

(The .local draft had SIGABRT/ABRT which is wrong: ABRT is not a valid SI_CODE token. And SIGTRAP/TRAP Trace should be SIGTRAP/TRAP_TRACE. And SIGVTALRM/VTALRM does not exist as a per-signal SI_CODE.)

#Index-friendly facets (high-value)

These fields are the first-pass slicers for queries. Always include at least 1-2 of these in selections before falling back to FTS:

• AE_AGENT_VERSION -- regression / fix-detection.
• AE_AGENT_HEALTH -- crash class.
• AE_EXIT_CAUSE -- exit class.
• AE_FATAL_SIGNAL_CODE -- signal type.
• AE_FATAL_FUNCTION -- localize to a function.
• AE_FATAL_FILENAME -- localize to a file.
• AE_HOST_ARCHITECTURE -- arch-specific bugs.
• AE_OS_FAMILY -- distro-specific bugs.
• AE_AGENT_PROFILE_0 (and _1, _2) -- parent / child / iot.
• AE_AGENT_KUBERNETES -- k8s-specific.
• AE_AGENT_INSTALL_TYPE -- packaging-specific.

#Privacy-sensitive fields

Treat these as identifying. The redact-events.sh opt-in filter masks them when sharing:

• AE_AGENT_ID (machine GUID).
• AE_HOST_ID (OS machine-id).
• AE_AGENT_NODE_ID, AE_AGENT_CLAIM_ID (Cloud identifiers).
• AE_HOST_BOOT_ID, AE_AGENT_EPHEMERAL_ID.
• AE_HW_SYS_UUID.
• DMI fields (AE_HW_*) when correlated with serial-equivalent identifiers.

(Privacy-sensitive serials and asset_tags are already commented out at the producer side and never reach the journal -- status-file.c:275-276, :294-295, :304-305.)

#What is NOT in the journal

Several producer fields are intentionally redacted at the producer side (commented out in dsf_json_hw):

• hw.sys.serial, hw.sys.asset_tag
• hw.board.serial, hw.board.asset_tag
• hw.chassis.serial, hw.chassis.asset_tag

There is no agent.happiness field in the producer source at any version. The .local draft mentioned it -- the field has never existed.

#Stack trace addresses are NOT anonymized in the journal

status-file-dedup.c:26-36 zeroes out hex addresses ONLY when computing the dedup hash. The journal-emitted AE_FATAL_STACK_TRACE retains real addresses. Useful for bug investigation; sensitive when sharing externally.