version bumped to 0.4.0, docs update, open api spec

Author: SuperioOne

CHANGELOG

@@ -1,3 +1,37 @@
+# v0.4.0 - October 12, 2024
+
+## Changed
+
+- Bump `apexcharts` version from `3.50.0` to `3.53.0`.
+- Bump `htmx` version from `2.0.1` to `2.0.3`.
+- Bump server crate versions to latest.
+- Generalized HTTP error responses based on RFC9457/RFC7807.
+- JSON API returns HTTP 5xx error codes with problem details when daemon connection is lost.
+- Logger redacts `upsd_pass` config when server starts with `LOG_LEVEL=DEBUG`
+- Logging style is improved in general. No more `{:?}` dumps.
+- Remaining battery runtime shown in yMdhms date-time format instead of raw seconds.
+- UPS table elements are sorted by UPS name.
+- Voltage texts are vertically centered on the UI.
+- `health` probe returns HTTP 500 status code when daemon connection is lost.
+- `health` probe returns application/json response. Response contains basic info like last sync time, daemon status, and daemon host.
+
+## Added
+
+- Display UPS manufacturer and model information on UPS page.
+- `readiness` probe for Kubernetes deployments.
+- upsd_client logs upsd protocol messages when `LOG_LEVEL` is set to `TRACE`.
+- OpenAPI specification is added to docs.
+
+## Fixed
+
+- A single failing UPS driver distrupts other UPS devices' states.
+- Probe endpoints don't support CORS.
+- Server doesn't delete failing/disconnected UPS devices from both HTML and JSON responses.
+- Texts get overlap in table elements when screen size is too small.
+- UPS info page displays incorrect power unit for the apparent power values (ups.power).
+- On certain light themes, text colors blend with background and becomes unreadable.
+- upsd_client uses wrong error types when list commands fail.
+
 # v0.3.2 - September 29, 2024
 
 ## Fixed

README.md

@@ -2,7 +2,15 @@
 
 Light weight web interface for [Network UPS Tools](https://networkupstools.org/).
 
-**Quickstart:**
+<div style="width: 100%; display: flex; justify-content: space-between; flex-direction: row; gap: 1rem;">
+
+![DetailImage](docs/images/details.webp)
+
+![ListImage](docs/images/home.webp)
+
+</div>
+
+## Quickstart
 
 ```shell
 docker run -e UPSD_ADDR=10.0.0.1 -e UPSD_USER=test -e UPSD_PASS=strongpass -p 9000:9000 ghcr.io/superioone/nut_webgui:latest
@@ -19,13 +27,18 @@ docker run -e UPSD_ADDR=10.0.0.1 -e UPSD_USER=test -e UPSD_PASS=strongpass -p 90
 > In order to run `INSTCMD`, make sure the configured user has proper instcmds granted at `upsd.users`. See
 > man([upsd.users](https://networkupstools.org/docs/man/upsd.users.html)).
 
+## Examples
+
+- [Kubernetes with EndpointSlice](docs/kubernetes_example.md)
+- [Docker compose](docs/docker_compose.md)
+
 ## CPU architecture support
 
 | Arch         | Test Hardware           | Notes                                                                                    |
 |--------------|-------------------------|------------------------------------------------------------------------------------------|
-| amd64        | AM4                     | Works across all amd64 platforms.                                                        |
-| amd64-v3     | AM4                     | Snake oil level optimizations with AVX. It mostly reduces response compression overhead. |
-| amd64-v4     | Not Tested              | Snake oil level optimizations with AVX-512.                                              |
+| amd64        | AM4 CPU                 | Works across all amd64 platforms.                                                        |
+| amd64-v3     | AM4 CPU                 | Snake oil level optimizations with AVX. It mostly reduces response compression overhead. |
+| amd64-v4     | Intel® SDE              | Snake oil level optimizations with AVX-512.                                              |
 | arm64        | Raspberry Pi 4 Model B  |                                                                                          |
 | armv7        | Qemu emulation          |                                                                                          |
 | armv6        | Qemu emulation          |                                                                                          |
@@ -34,11 +47,6 @@ docker run -e UPSD_ADDR=10.0.0.1 -e UPSD_USER=test -e UPSD_PASS=strongpass -p 90
 > amd64 v3 and v4 variants require certain CPU feature flags to run. If you are a crackhead min-max enjoyer (like me), you can use `nut_webgui:latest-amd64-v3` and `nut_webgui:latest-amd64-v4` images.
 > See [x86-64 Microarchitecture levels](https://en.wikipedia.org/wiki/X86-64#Microarchitecture_levels) for more details.
 
-![DetailImage](docs/images/details.webp)
-
-![ListImage](docs/images/list.webp)
-
-![INSTCMDImage](docs/images/inst_cmd.webp)
 
 ## Command-Line arguments
 
@@ -67,18 +75,15 @@ docker run -e UPSD_ADDR=10.0.0.1 -e UPSD_USER=test -e UPSD_PASS=strongpass -p 90
 
 ## JSON data API
 
-A simple JSON-based API is available for integration and automation purposes. For usage details
-see [JSON data API page.](./docs/json_api.md)
-
-## Probes
+A simple JSON-based API is available for integration and automation purposes.
 
-The server offers a basic health probe at the `/probes/health` endpoint. This can be used as liveness or readiness
-checks in your systems.
+OpenAPI 3.1.0 specification files: [json minimized](./docs/openapi3_spec.json) | [yaml](./docs/openapi3_spec.yaml)
 
-## Examples
+## Probes
 
-- [Kubernetes with EndpointSlice](docs/kubernetes_example.md)
-- [Docker compose](docs/docker_compose.md)
+nut_webgui has basic probe endpoints to check server health and readiness:
+- `/probes/health`
+- `/probes/readiness`
 
 ## Building from source and debugging
 

client/package.json

@@ -5,7 +5,7 @@
   },
   "private": true,
   "name": "nut_webgui-client",
-  "version": "0.3.2",
+  "version": "0.4.0",
   "description": "Bundles all client side web components and dependencies.",
   "scripts": {
     "dev-js": "esbuild ./src/index.js --bundle --outdir=./dist/debug --target=firefox98,chrome86,safari15 --format=iife --watch=forever",

docs/images/details.webp

@@ -92,21 +92,31 @@ spec:
               value: "20"
           resources:
             limits:
-              memory: "128Mi"
+              memory: "64Mi"
             requests:
-              memory: "128Mi"
+              memory: "32Mi" # You can go low as 4Mi. As example, single pod uses 2Mi memory for a user with 12 UPS devices.
           ports:
             - containerPort: 9000
+          startupProbe:
+            httpGet:
+              path: /probes/readiness
+              port: 9000
+            initialDelaySeconds: 5
+            failureThreshold: 15
+            periodSeconds: 10
           livenessProbe:
             httpGet:
               path: /probes/health
               port: 9000
             initialDelaySeconds: 5
+            failureThreshold: 3
             periodSeconds: 30
           readinessProbe:
             httpGet:
-              path: /probes/health
+              path: /probes/readiness
               port: 9000
-
+            initialDelaySeconds: 5
+            failureThreshold: 3
+            periodSeconds: 30
 ```