Sd 22337 monarch fine grained control real time telemetry selection feature by stevencarpenter102 · Pull Request #19 · FireflySpace/fprime

stevencarpenter102 · 2026-08-19T23:07:41Z


*Related Issue(s)*	`SD-22337-monarch-fine-grained-control-real-time-telemetry-selection-feature`
*Has Unit Tests (y/n)*	y
*Documentation Included (y/n)*	n
*Generative AI was used in this contribution (y/n)*	y

Change Description

support per-packet configuration, write out the per-packet overrides to persistent storage via a configOut port to a component of your choosing which manages the storage; at boot-up pull in config written to persistent storage via configIn port

Rationale

facilitates "finer-grain" control of packet telemetry configuration (rather than the more "coarse" control at the TlmPacketizer's group-level); this "finer-grain" per-packet control is being requested from some of FF's missions

Testing/Review Recommendations

testing focused on use of the 'bg2_flight' deployment w/ Linux + mpac builds on local laptop + Racksat hardware
confirmed w/ fsw binary running that ENABLE_PACKET() and FORCE_PACKET() influence run-time behavior as expected

Future Work

n/a

AI Usage (see policy)

understood

…ASSERTs

This reverts commit f48314f.

… NAK processed logic

…that hook into the tlmpacketizer

…fied time not nanosec)

… fail build until this commit's change

… input before), as per recommendation of Michael Starch

stevencarpenter102 · 2026-08-25T18:57:07Z

+
+    @ Maximum number of per-packet config entries carried in a single push batch.
+    @ Sized so one fully-serialized batch stays well within FW_COM_BUFFER_MAX_SIZE.
+    constant TLM_PACKET_CONFIG_BATCH_MAX = 32


note this "batched config write" feature is required b/c of the FW_COM_BUFFER_MAX_SIZE (~2k) limitation means we can only send out the configOut port so many per-packet override configs serialized at a time...

This should be moved to TlmPackizerCfg.fpp

stevencarpenter102 · 2026-08-25T19:19:12Z

+    this->invoke_to_TlmRecv(0, 100, ts, buff);        // second channel
+    buff.resetSer();
+    (void)buff.serializeFrom(static_cast<U8>(14));
+    this->invoke_to_TlmRecv(0, 333, ts, buff);        // third channel


skipping the fourth channel is intentionally...following same pattern of skipping found in many other unit tests throughout this file

stevencarpenter102 · 2026-08-25T21:15:29Z

    return outIndex;
 }

+Svc::PacketConfig TlmPacketizer::effectiveConfig(FwIndexType section, FwChanIdType pkt, FwChanIdType group) const {


"effective" in the sense of prioritize the per-packet override...and if there's none -> then use the group config of which the packet is a part

stevencarpenter102 · 2026-08-25T21:25:50Z

+                                          rateLogic: RateLogic        @< Rate logic
+                                          minDelta: U32               @< Minimum Sched ticks between sends (ON_CHANGE_MIN logic)
+                                          maxDelta: U32               @< Maximum Sched ticks between sends (EVERY_MAX logic)
+                                        ) \


the three functions above mirror the same functionality available for the "group"-level...now we're exposes the "per-packet" more 'fine-grain' level

stevencarpenter102 · 2026-08-25T21:26:17Z

    telemetry SectionEnabled: SectionEnabled id 1

+    @ Effective per-packet configuration, emitted in response to GET_PACKET_CONFIG
+    telemetry QueriedPacketConfig: PacketConfigEntry id 2


notice the comment here...esp useful for debugging

…des after reboot

stevencarpenter102 · 2026-08-25T22:48:58Z

+
+    @ Per-packet configuration reload from the persistence storage component
+    @ That storage component pushes it's config into this port when commanded to do so (usually at boot-up)
+    async input port configIn: TlmPacketConfigUpdate


this was motivated by the use-case...

somehow the TlmPacketizer needs to get the persisted packet-overrides after a reboot...this port is the way you get this config in here

stevencarpenter102 · 2026-08-25T22:50:30Z

+    async command GET_PACKET_CONFIG(
+                                    packetId: U32               @< Packet identifier
+                                    section: TelemetrySection   @< Section to query
+                                  ) \


after sending this command check the QueriedPacketConfig telemetry channel for the latest packet config asked for!

stevencarpenter102 · 2026-08-25T22:51:12Z

+    //! CONFIGURE_PACKET_RATES commands. When m_packetOverridden is set, this value wins over
+    //! the group-derived policy for that packet/section. This is the per-packet control layer;
+    //! group config remains the base for non-overridden packets. Each change is mirrored out
+    //! configOut to the passive TlmPacketConfig for persistence.


packet override >> group policy

stevencarpenter102 · 2026-08-25T22:52:27Z

+      rateLogic: RateLogic      @< Rate logic configuration
+      min: U32                  @< Minimum Sched ticks between sends when using ON_CHANGE_MIN logic
+      max: U32                  @< Maximum Sched ticks between sends when using EVERY_MAX logic
+    }


PacketConfig mirrors the struct GroupConfig

Brian-Campuzano

Recommend updating the SDD to keep track of how packet level overrides interact with the existing "knobs". Also need to update the component port interface definition.

Brian-Campuzano · 2026-09-03T13:50:23Z

+
+    @ Maximum number of per-packet config entries carried in a single push batch.
+    @ Sized so one fully-serialized batch stays well within FW_COM_BUFFER_MAX_SIZE.
+    constant TLM_PACKET_CONFIG_BATCH_MAX = 32


This should be moved to TlmPackizerCfg.fpp

Brian-Campuzano · 2026-09-03T13:52:56Z

+    }
+    s = static_cast<FwSizeType>(section.e);
+    if (not this->m_packetOverridden[s][pkt]) {
+        this->m_packetOverride[s][pkt] = TlmPacketizer::defaultPacketConfig();


Should this be overriding with the effectiveConfig() instead of the defaultPacketConfig()?

Brian-Campuzano · 2026-09-03T13:54:06Z

+    if (this->m_packetOverridden[s][pkt]) {
+        return this->m_packetOverride[s][pkt];


Once a packet has been overridden, how is it cleared? Do we need a new command?

Brian-Campuzano · 2026-09-03T13:56:27Z

+    // Not overridden: derive from the group policy for this packet's level (legacy behavior).
+    const TlmPacketizer_GroupConfig& gc = this->m_groupConfigs[s][group];
+    Svc::PacketConfig eff;
+    eff.set_enabled(gc.get_enabled());
+    eff.set_forceEnabled(gc.get_forceEnabled());
+    eff.set_rateLogic(gc.get_rateLogic());
+    eff.set_min(gc.get_min());
+    eff.set_max(gc.get_max());
+    return eff;


This and the return on 652 is building/copying a new PacketConfig and returning by value every evaluation cycle. Consider refactoring to return a reference a performance improvement.

Brian-Campuzano · 2026-09-03T13:57:09Z

+void TlmPacketizer ::configIn_handler(FwIndexType portNum, FwSizeType count, const Svc::PacketConfigBatch& batch) {
+    // load overrides from an external component's persistant storage of the overrides (intended to be used after a reboot)
+    const FwSizeType cap = static_cast<FwSizeType>(Svc::PacketConfigBatch::SIZE);
+    const FwSizeType n = (count < cap) ? count : cap;


Consider a warning EVR if the cap is exceeded.

Brian-Campuzano added 30 commits July 6, 2026 14:55

Added CfdpManager skeleton

e7d9b28

Copy PDU packet layer as-is

a8d20de

Bring in other header dependencies

676c595

Updates to get the packet layer to compile

7b7b248

Bring in CFDP engine files

a8417b3

Bring in addititional CF depedencies and refactor CF asserts into FW_…

44d0660

…ASSERTs

c -> c++ updates

a81d937

c to c++ conversion checkpoint

2194258

Os directory checkpoint

89fcab3

Add custom timer soloution to replace CFE timer implementation

ffaccb8

Os File stub completion

e5542e5

Update CRC

cec7325

Added buffer management and removed pipes/semaphores

1111a33

Beggining of buffer refactor

8364b49

Buffer management checkpoint

fa2548c

Brought in cf_utils for list traversals

4ef83b9

Convert CfdpEngine into a class

36423d7

Revert "Convert CfdpEngine into a class"

5afba79

This reverts commit f48314f.

Additional CFE OS stubs

d733f34

Fixed container pointer arithmatic for C++

e5ac3b4

Completed refactor of cf_utils

1cbf9e5

Beginning of adding parameters

68ed031

Refactor CF_CFDP_S_CheckAndRespondNak to separate out error logic and…

07a2bee

… NAK processed logic

CFDP send checkpoint

eb29119

Completed CFDP send updates

385f5e7

Completed CFDP receive refactor

b2e2b44

CFDP refactor checkpoint

2886c73

Playback directory checkpoint

85976a8

CfdpManager compiles!

64ddc57

Added PDU receive data paths/hooks

d1d6ee0

pepepr08 and others added 12 commits July 6, 2026 14:57

Increase PDU buffer size to include Descriptor

a5a4b77

Add helper function for sending PDUs

310525f

Move PDU Type Enum to FPP

e397fef

Assert to nullptr instead of just NULL

8a72a58

phase 1 of implementing "per packet" enable+rate+etc control "knobs" …

82515ba

…that hook into the tlmpacketizer

Make the cmake dictionary flow track inputs and outputs

3d41df1

Metadata clear our root dir (avoid ninja nothing to do spam)

3509b8e

touch dict after install (cmake natively writes whole seconds to modi…

ff83a11

…fied time not nanosec)

phase 2 complete - introduced per-packet tlmpacktizer config overrides

f3be4d3

merge origin/monarch attempt! this was quite messy, lol

9ad6b8e

revert mistaken from previous commit's merge attempt

43b8057

last touches from last commit's attempt to fix

346181b

pepepr08 force-pushed the monarch branch from ff83a11 to 78fc4b0 Compare August 24, 2026 16:34

stevencarpenter102 added 3 commits August 24, 2026 13:05

she builds now! before after origin/dev merge in monarch...this would…

18d87a1

… fail build until this commit's change

merge in latest origin/monarch

5212390

switched direction of TlmPacketizer's port (now it's an output, was a…

8a87c59

… input before), as per recommendation of Michael Starch

stevencarpenter102 self-assigned this Aug 25, 2026

stevencarpenter102 commented Aug 25, 2026

View reviewed changes

introducing the configIn port in order to get persisted packet overri…

b9d5b82

…des after reboot

stevencarpenter102 commented Aug 25, 2026

View reviewed changes

stevencarpenter102 requested review from Brian-Campuzano and Lex-ari August 25, 2026 22:52

Brian-Campuzano requested changes Sep 3, 2026

View reviewed changes

		if (this->m_packetOverridden[s][pkt]) {
		return this->m_packetOverride[s][pkt];

Uh oh!

Conversation

stevencarpenter102 commented Aug 19, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Change Description

Rationale

Testing/Review Recommendations

Future Work

AI Usage (see policy)

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

stevencarpenter102 Aug 25, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Brian-Campuzano left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

stevencarpenter102 commented Aug 19, 2026 •

edited

Loading

stevencarpenter102 Aug 25, 2026 •

edited

Loading