4 releases (2 breaking)
| 0.3.1 | Sep 23, 2023 |
|---|---|
| 0.3.0 | Sep 22, 2023 |
| 0.2.1 |
|
| 0.2.0 | Aug 25, 2023 |
| 0.1.0 | Aug 14, 2023 |
#407 in Debugging
215KB
863 lines
seella
A tool for visualising the traces emitted by ScyllaDB.
Generates waterfall charts and a tree diagram of system_tracing.events in your terminal.
Basic invocation:
$ seella "74ff67c0-397b-11ee-8ca4-9688db6cc0f1"
Session ID: 74ff67c0-397b-11ee-8ca4-9688db6cc0f1
2023-08-13T01:48:10.172+00:00
172.17.0.2 (anonymous) -> 172.17.0.2
Request Size: 84
Response Size: 20
Execute CQL3 query
"{'consistency_level': 'ONE', 'page_size': '100', 'query': 'INSERT INTO k.t (pk, t, v, s) VALUES (0, 1, ''val1'', ''static1'');', 'serial_consistency_level': 'SERIAL', 'user_timestamp': '1691891290172041'}"
waterfall chart dur node activity
1 [█─────┤ ] 0 172.17.0.2 ├┬─ Parsing a statement
2 [█ ] 5 172.17.0.3 │├─ Message received from /172.17.0.2
3 [███ ] 55 172.17.0.3 │├─ Sending mutation_done to /172.17.0.2
4 [ ████ ] 58 172.17.0.3 │├─ Mutation handling is done
5 [ █ ] 27 172.17.0.2 ├── Processing a statement
6 [ ███ ] 45 172.17.0.2 ├── Creating write handler for token: -3485513579396041028 natural: {172.17.0.3} pending: {}
7 [ ███ ] 46 172.17.0.2 ├── Creating write handler with live: {172.17.0.3} dead: {}
8 [ ███ ] 49 172.17.0.2 ├── Sending a mutation to /172.17.0.3
9 [ ████████████████████ ] 332 172.17.0.2 ├── Got a response from /172.17.0.3
10 [ █████████████████████ ] 333 172.17.0.2 ├── Delay decision due to throttling: do not delay, resuming now
11 [ ████████████████████ ] 339 172.17.0.2 ├── Mutation successfully completed
12 [ ██████████████████████] 344 172.17.0.2 ├── Done processing - preparing a result
Or one with more options:
$ seella -w 50 --show-event-id --show-span-ids --show-thread --max-activity-width 50 "74ff67c0-397b-11ee-8ca4-9688db6cc0f1"
Session ID: 74ff67c0-397b-11ee-8ca4-9688db6cc0f1
2023-08-13T01:48:10.172+00:00
172.17.0.2 (anonymous) -> 172.17.0.2
Request Size: 84
Response Size: 20
Execute CQL3 query
"{'consistency_level': 'ONE', 'page_size': '100', 'query': 'INSERT INTO k.t (pk, t, v, s) VALUES (0, 1, ''val1'', ''static1'');', 'serial_consistency_level': 'SERIAL', 'user_timestamp': '1691891290172041'}"
waterfall chart dur node activity event id span id parent span id thread name
1 [█─┤ ] 0 172.17.0.2 ├┬─ Parsing a statement 74ff70c8-397b-11ee-8ca4-9688db6cc0f1 153249663699531 0 shard 0
2 [█ ] 5 172.17.0.3 │├─ Message received from /172.17.0.2 74ff75b2-397b-11ee-a288-20cc230d8ac0 343569500103777 153249663699531 shard 4
3 [█ ] 55 172.17.0.3 │├─ Sending mutation_done to /172.17.0.2 74ff779d-397b-11ee-a288-20cc230d8ac0 343569500103777 153249663699531 shard 4
4 [ ██ ] 58 172.17.0.3 │├─ Mutation handling is done 74ff77c2-397b-11ee-a288-20cc230d8ac0 343569500103777 153249663699531 shard 4
5 [ █ ] 27 172.17.0.2 ├── Processing a statement 74ff71dc-397b-11ee-8ca4-9688db6cc0f1 153249663699531 0 shard 0
6 [ █ ] 45 172.17.0.2 ├── Creating write handler for token: -348551357939604 74ff728a-397b-11ee-8ca4-9688db6cc0f1 153249663699531 0 shard 0
7 [ ██ ] 46 172.17.0.2 ├── Creating write handler with live: {172.17.0.3} dea 74ff7296-397b-11ee-8ca4-9688db6cc0f1 153249663699531 0 shard 0
8 [ █ ] 49 172.17.0.2 ├── Sending a mutation to /172.17.0.3 74ff72b1-397b-11ee-8ca4-9688db6cc0f1 153249663699531 0 shard 0
9 [ ██████████ ] 332 172.17.0.2 ├── Got a response from /172.17.0.3 74ff7dc2-397b-11ee-8ca4-9688db6cc0f1 153249663699531 0 shard 0
10 [ ███████████ ] 333 172.17.0.2 ├── Delay decision due to throttling: do not delay, re 74ff7dcb-397b-11ee-8ca4-9688db6cc0f1 153249663699531 0 shard 0
11 [ ██████████ ] 339 172.17.0.2 ├── Mutation successfully completed 74ff7e09-397b-11ee-8ca4-9688db6cc0f1 153249663699531 0 shard 0
12 [ ███████████] 344 172.17.0.2 ├── Done processing - preparing a result 74ff7e3a-397b-11ee-8ca4-9688db6cc0f1 153249663699531 0 shard 0
Usage
See seella --help for all the options:
$ seella --help
A tool for visualising the traces emitted by ScyllaDB
Usage: seella [OPTIONS] <COMMAND>
Commands:
csv Use a pair of CSVs as a data source
db Use a live database as a data source
help Print this message or the help of the given subcommand(s)
Options:
-w, --waterfall-width <WATERFALL_WIDTH>
The width of the waterfall chart [default: 100]
-d, --duration-format <DURATION_FORMAT>
Whether to generate span durations in milliseconds or microseconds [default: micros] [possible values: millis, micros]
--min-duration-width <MIN_DURATION_WIDTH>
Minimum print width for the duration field, remaining will be filled with spaces [default: 6]
--max-activity-width <MAX_ACTIVITY_WIDTH>
Maximum print width for the activity field, remaining will be truncated [default: 300]
--show-event-id
Whether to show the event uuid
--show-span-ids
Whether to show the span ids
--show-thread
Whether to show the thread name
-h, --help
Print help
-V, --version
Print version
$ seella csv --help
Use a pair of CSVs as a data source
Usage: seella csv [OPTIONS] <SESSION_ID>
Arguments:
<SESSION_ID> The session id to be visualised
Options:
-s, --sessions-path <SESSIONS_PATH> Path to the CSV containing the sessions data. Any string that can be coerced into a PathBuf [default: sessions.csv]
-e, --events-path <EVENTS_PATH> Path to the CSV containing the events data. Any string that can be coerced into a PathBuf [default: events.csv]
-h, --help Print help
$ seella db --help
Use a live database as a data source
Usage: seella db [OPTIONS] <SESSION_ID>
Arguments:
<SESSION_ID> The session id to be visualised
Options:
-a, --addr <ADDR> Socket Address (IP address and port) for the database connection. See [std::net::SocketAddr::from_str] [default: 127.0.0.1:9042]
-h, --help Print help
Samples
Run the following:
make up
make cqlsh
And then within that cqlsh session:
TRACING ON;
CREATE KEYSPACE k
WITH REPLICATION = {
'class' : 'NetworkTopologyStrategy',
'datacenter1' : 1
};
CREATE TABLE k.t (
pk int,
t int,
v text,
s text,
PRIMARY KEY (pk, t)
);
INSERT INTO k.t (pk, t, v, s) VALUES (0, 0, 'val0', 'static0');
COPY system_traces.sessions TO '/data/sessions.csv' WITH HEADER = TRUE;
COPY system_traces.events TO '/data/events.csv' WITH HEADER = TRUE;
You will now have a sessions.csv and events.csv in your local directory that you can experiment with.
Run make down when you're done to shut down the cluster.
Dependencies
~12–22MB
~300K SLoC