Event Feed » History » Version 2
laforge, 06/22/2018 05:03 PM
formatting
| 1 | 1 | laforge | {{>toc}} |
|---|---|---|---|
| 2 | 2 | laforge | |
| 3 | 1 | laforge | h1. Event Feed |
| 4 | |||
| 5 | General idea: Feed of events pushed out from [each] Osmocom program |
||
| 6 | |||
| 7 | h2. Events |
||
| 8 | |||
| 9 | h3. FSM creation |
||
| 10 | |||
| 11 | * timestamp |
||
| 12 | * source file (string) |
||
| 13 | * source line (u32) |
||
| 14 | * ptr (u32/u64) |
||
| 15 | * fsm class (string) |
||
| 16 | * ID (string) |
||
| 17 | * parent ptr (if any) |
||
| 18 | |||
| 19 | h3. FSM event |
||
| 20 | |||
| 21 | * timestamp |
||
| 22 | * source file (string) |
||
| 23 | * source line (u32) |
||
| 24 | * ptr (u32/64) |
||
| 25 | * event number (u8/32) |
||
| 26 | * state number (u8/32) |
||
| 27 | * decoded/interpreted event payload? tlv? |
||
| 28 | |||
| 29 | h3. FSM state change |
||
| 30 | |||
| 31 | * timestamp |
||
| 32 | * source file (string) |
||
| 33 | * source line (u32) |
||
| 34 | * ptr (u32/64) |
||
| 35 | * old state number (u8/32) |
||
| 36 | * new state number (u8/32) |
||
| 37 | * timer number (int) |
||
| 38 | * timer timeout (u32) |
||
| 39 | ** optional. old timer could also remain running! |
||
| 40 | |||
| 41 | h3. FSM termination |
||
| 42 | |||
| 43 | * timestamp |
||
| 44 | * source file (string) |
||
| 45 | * source line (u32) |
||
| 46 | * ptr (u32/u64) |
||
| 47 | * termination cause |
||
| 48 | |||
| 49 | h3. FSM rename |
||
| 50 | |||
| 51 | * timestamp |
||
| 52 | * source file (string) |
||
| 53 | * source line (u32) |
||
| 54 | * ptr (u32/u64) |
||
| 55 | * old id (string) ? |
||
| 56 | * new id (string) |
||
| 57 | |||
| 58 | h3. FSM reparent (change or unlink) |
||
| 59 | |||
| 60 | * timestamp |
||
| 61 | * ptr (u32/u64) |
||
| 62 | * old parent ptr (u32/u64) |
||
| 63 | * new parent ptr (u32/u64) |
||
| 64 | |||
| 65 | h3. non-FSM events |
||
| 66 | |||
| 67 | * measurement reports (bts/bsc) |
||
| 68 | * "jolly fsm" like LAPDm, CC FSMs -> maybe convert them? |
||
| 69 | * OsmoPCU: TBF / resoure block allocations per frame number |
||
| 70 | ** probably needs some quite different handling? |
||
| 71 | |||
| 72 | |||
| 73 | h2. filtering |
||
| 74 | |||
| 75 | * ability to enable/disable generation of events per FSM class [or instance?] |
||
| 76 | * ability to do retroactive filtering in the visualization tool |
||
| 77 | |||
| 78 | h2. format / encoding |
||
| 79 | |||
| 80 | * don't generate JSON directly, let's avoid all the string encoding |
||
| 81 | * use CTRL or other interface to query string tables of FSM events and FSM states at runtime |
||
| 82 | * external program can use string tables + binary protocol to generate JSON feed for browser |
||
| 83 | * external program can collect/aggregate feeds from multiple osmocom programs |
||
| 84 | |||
| 85 | h2. visual representation |
||
| 86 | |||
| 87 | * list of instances per FSM class |
||
| 88 | ** sort it by "most recent", "ID" or "will expire soonest" |
||
| 89 | * history / time line for each FSM instance |
||
| 90 | * relationship of parent/child FSMs, maybe some kind of tree structure? |
||
| 91 | * show FSMs of different osmocom network elements in one view, i.e. BTS+BSC+MSC |
||
| 92 | |||
| 93 | |||
| 94 | h2. misc ideas |
||
| 95 | |||
| 96 | * pass out ptr/addr as u32/u64 to uniquely identify FSM |
||
| 97 | ** string ID would then only be needed during creation and rename |
||
| 98 | ** FSM class would then only be needed during creation |
||
| 99 | ** u32/u64 would be unique identifier for all FSM instances inside one osmo-* program |
||
| 100 | * generate "FSM create" events with current state at start-up (optionally?) |
