You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: bench/tx-generator/CHANGELOG.md
+7
Original file line number
Diff line number
Diff line change
@@ -1,10 +1,17 @@
1
1
# ChangeLog
2
2
3
+
## 2.15 -- Mar 2025
4
+
5
+
* A new cabal flag `withplutuslib` is added, enabling import and re-compilation of Plutus scripts from `plutus-scripts-bench` - default: false; use for dev/test of new benchmarks only.
6
+
* Port `ProtocolParameters` type and typeclass instances from `cardano-api` into new module `Cardano.Api.Internal`, removing dependency on the deprecated API type.
7
+
* Bump for Node 10.3
8
+
3
9
## 2.14.2 -- Oct 2024
4
10
5
11
* Bump for Node 10
6
12
7
13
## 2.14.1 -- June 2024
14
+
8
15
* A new NixSvcOptions field is introduced: `_nix_keepalive`
9
16
and it's propagated down to the `kaClient` that does keepalives.
Copy file name to clipboardExpand all lines: cardano-tracer/CHANGELOG.md
+9
Original file line number
Diff line number
Diff line change
@@ -1,5 +1,14 @@
1
1
# ChangeLog
2
2
3
+
## 0.3.2 (March 2025)
4
+
5
+
* When requesting forwarded metrics, ask for delta to previous request only. New config option `ekgRequestFull` defaults to `false`; set to `true` to revert this behavior.
6
+
* Add up-to-date symlink for each log rotation, pointing to the most recent file (`node.json` or `node.log`, depending on the log format).
7
+
* Fix: CLI option `--min-log-severity` is now observed correctly.
8
+
* Add `metricsNoSuffix` to configuration, which when set removes suffixes like `_int` from metrics names, making them more similar to those in the old system (optional boolean; default: `false`).
9
+
* Remove `metricsComp` metric name remapping from configuration.
10
+
* Updated to use `ekg-forward-0.9`.
11
+
3
12
## 0.3.1 (January 22, 2025)
4
13
5
14
* Fix race condition when querying `NodeInfo` data point, occasionally resulting in fallback `NodeName`s instead of proper ones.
Copy file name to clipboardExpand all lines: cardano-tracer/docs/cardano-tracer.md
+1-1
Original file line number
Diff line number
Diff line change
@@ -296,7 +296,7 @@ In this example, `subdirForNode0` is a subdirectory containing log files with it
296
296
297
297
The field `logMode` specifies logging mode. There are two possible modes: `FileMode` and `JournalMode`. `FileMode` is for storing logs to the files, `JournalMode` is for storing them in `systemd`'s journal. Please note that if you choose `JournalMode`, the field `logRoot` will be ignored.
298
298
299
-
The field `logFormat` specifies the format of logs. There are two possible modes: `ForMachine` and `ForHuman`. `ForMachine` is for JSON format, `ForHuman` is for human-friendly text format.
299
+
The field `logFormat` specifies the format of logs. There are two possible modes: `ForMachine` and `ForHuman`. `ForMachine` is for JSON format, `ForHuman` is for human-friendly text format. The latter also implies the file endings change to `.log`, and the symbolic link is named `node.log`.
300
300
301
301
Please note that `logging` field accepts the list, so you can specify more than one logging section. For example, for both log formats:
Copy file name to clipboardExpand all lines: trace-dispatcher/CHANGELOG.md
+9
Original file line number
Diff line number
Diff line change
@@ -1,5 +1,14 @@
1
1
# Revision history for trace-dispatcher
2
2
3
+
## 2.9 -- Mar 2025
4
+
5
+
* New `PrometheusSimple` backend which runs a simple TCP server for direct exposition of metrics, without forwarding.
6
+
* New `maxReconnectDelay` config option in `TraceOptionForwarder`: Specifies maximum delay (seconds) between (re-)connection attempts of a forwarder (default: 60s).
7
+
* Optimize data sharing in formatters.
8
+
* Remove unused optional namespace prefix argument from formatters.
Copy file name to clipboardExpand all lines: trace-dispatcher/doc/trace-dispatcher.md
+20-14
Original file line number
Diff line number
Diff line change
@@ -305,8 +305,19 @@ The configurability of dispatchers provided by this library relies on:
305
305
The usual form to provide a configuration is via a configuration file, wich can be inJSONorYAML format.The options that
306
306
can be given based on a namespace are:`severity`, `detail`, `backends`and`limiter`.
307
307
308
-
Backends can be a combination of `Forwarder`, `EKGBackend`, and
309
-
one of `StdoutMachineFormat`, `tdout HumanFormatColoured` and `StdoutHumanFormatUncoloured`.
308
+
Backends can be a combination of `Forwarder`, `EKGBackend`, `PrometheusSimple [suffix|nosuffix] [bindhost] <port>` and
309
+
one of `Stdout MachineFormat`, `Stdout HumanFormatColoured` and `Stdout HumanFormatUncoloured`.
310
+
311
+
The connection for the `Forwarder` backend is provided on the application command line. It is a socket path over which applications like `cardano-node` connect with `cardano-tracer`. `--tracer-socket-path-connect /path/to/forward.sock` sets
312
+
the backends's role to `Initiator`, whereas `--tracer-socket-path-accept /path/to/forward.sock` sets it to `Responder`. Except for debugging purposes, the former should be chosen: the application takes the `Initiator` role, and `cardano-tracer` is
313
+
in the `Responder` role, which means setting its network `tag` to `AcceptAt` in its config (see there).
314
+
315
+
The `PrometheusSimple` backend provides Prometheus metrics _directly from the process_, without forwarding. It always applies to all tracers globally, and should only be configured once.
316
+
Providing an available port number in the connection string is mandatory; this will bind to localhost only by default. By specifying a bind host, the metrics can be queried remotely, e.g. over IPv4 by
317
+
binding to `0.0.0.0`, or IPv6 by binding to `::`. Metrics will be available under the URL `/metrics`.
318
+
The `nosuffix` modifier removes suffixes like `_int` from metrics names, making them more similar to those in the old system; `suffix` is the implicit default and can be omitted.
319
+
320
+
*CAUTION*: Generally allowing remote queries of Prometheus metrics is risky and should only be done in an environment you control.
310
321
311
322
```yaml
312
323
# Use new tracing
@@ -320,6 +331,7 @@ TraceOptions:
320
331
- Stdout MachineFormat
321
332
- EKGBackend
322
333
- Forwarder
334
+
- 'PrometheusSimple :: 1234' # Prometheus metrics available over IPv6 (and localhost) on port 1234
323
335
324
336
ChainDB: # Show as well messages with severity Info for all ChainDB traces.
325
337
severity: Info
@@ -329,16 +341,13 @@ TraceOptions:
329
341
maxFrequency: 2.0
330
342
331
343
TraceOptionForwarder: # Configure the forwarder
332
-
address:
333
-
filePath:/tmp/forwarder.sock
334
-
mode:Initiator
335
344
maxReconnectDelay: 20
336
345
337
346
# Frequency of Peer messages set to two seconds
338
347
TraceOptionPeerFrequency: 2000
339
348
340
349
# Any metrics emittted will get this prefix
341
-
TraceOptionMetricsPrefix:"cardano.node"
350
+
TraceOptionMetricsPrefix: "cardano.node.metrics."
342
351
```
343
352
344
353
The same in JSON looks like this:
@@ -353,7 +362,8 @@ The same in JSON looks like this:
353
362
"backends": [
354
363
"Stdout MachineFormat",
355
364
"EKGBackend",
356
-
"Forwarder"
365
+
"Forwarder",
366
+
"PrometheusSimple :: 1234"
357
367
]
358
368
},
359
369
"ChainDB": {
@@ -365,14 +375,10 @@ The same in JSON looks like this:
0 commit comments