mirror of
https://github.com/Cacti/cacti.git
synced 2026-10-23 03:27:02 +09:00
docs: fix typos and grammar in api, locales, and service READMEs (#6690)
* docs: fix typos and grammar in api, locales, and service READMEs api/README.md: fix PHP version requirement (7.4→8.1), replace incorrect WSGI reference, fix formatting and punctuation. locales/README.md: fix grammar (The the→While the), correct spelling (on boarded→onboarded), fix typos (of if→or, if). service/README.md: fix grammar (entering→entered, it's→its), correct Debian user/group (www-run:www→www-data:www-data), fix wording (for as a future→as a future, over write→overwrite). No content changes; copy-editing only. Signed-off-by: Thomas Vincent <thomasvincent@gmail.com> * docs: add router script to PHP built-in server command PHP's built-in server doesn't route non-file paths to index.php by default, so Slim endpoints return 404. Adding public/index.php as the router script argument fixes this. Signed-off-by: Thomas Vincent <thomasvincent@gmail.com> --------- Signed-off-by: Thomas Vincent <thomasvincent@gmail.com>
This commit is contained in:
1 parent
31341e0706
commit
710a533f5a
3 files changed
+25
-19
No files matched your search
+10
-4
@@ -6,9 +6,15 @@ This project provides a RESTful API for accessing Cacti monitoring data using th
|
||||
|
||||
The API exposes endpoints to retrieve information about hosts, host templates, poller status, graph lists, Cacti status, boost status, database connectivity, and plugin thresholds. All responses are in JSON format.
|
||||
|
||||
THIS IS NOT PROD READY!
|
||||
**This API is not production ready.**
|
||||
|
||||
To run the API php -S 127.0.0.1:8080 -t public ( Which for Prod will be replaced with a WSGI server)
|
||||
To run the API for development:
|
||||
|
||||
```bash
|
||||
php -S 127.0.0.1:8080 -t public public/index.php
|
||||
```
|
||||
|
||||
For production, this should be served by a proper web server (Apache, Nginx, etc.).
|
||||
|
||||
## API Versioning
|
||||
|
||||
@@ -52,7 +58,7 @@ The API uses URL path versioning. All endpoints are prefixed with a version numb
|
||||
Checks database connectivity.
|
||||
|
||||
- `GET /v1/status/cacti_db_status`
|
||||
Returns some metrics of the Main cacti DB
|
||||
Returns metrics for the main Cacti database.
|
||||
|
||||
#### Plugin Endpoints
|
||||
- `GET /v1/plugin/thold/thresholds`
|
||||
@@ -69,7 +75,7 @@ The API uses URL path versioning. All endpoints are prefixed with a version numb
|
||||
|
||||
## Requirements
|
||||
|
||||
- PHP 7.4 or higher
|
||||
- PHP 8.1 or higher
|
||||
- Composer
|
||||
- Cacti database and configuration
|
||||
|
||||
|
||||
+5
-5
@@ -18,7 +18,7 @@ machine readable versions of the language translation per country and language.
|
||||
## Management Process
|
||||
|
||||
When making a pull request, if you are modifying human readable strings, it's
|
||||
important that the user modifying those strings to include an updated **cacti.pot**
|
||||
important that the user modifying those strings include an updated **cacti.pot**
|
||||
file in their pull requests. To update the **cacti.pot** file, you would simply
|
||||
follow the directions below:
|
||||
|
||||
@@ -32,7 +32,7 @@ This script will perform the following steps:
|
||||
- Update your copy of the various language translation files (*.po) per country and language
|
||||
- Update your copy of the various binary translation files (*.mo) per country and language
|
||||
|
||||
The the **build_gettext.sh** script updates both the *.po and *.mo files, those are not important
|
||||
While the **build_gettext.sh** script updates both the *.po and *.mo files, those are not important
|
||||
for your pull requests as Weblate updates those files automatically as soon as it detects
|
||||
the change in the **cacti.pot** file from your pull request.
|
||||
|
||||
@@ -51,10 +51,10 @@ into the Cacti GitHub repository, which will be a signal to the team. Once we s
|
||||
translation, we may have a few additional steps that need to be required on the Cacti site.
|
||||
Someone from the Cacti Development team will reach out to you with those instructions.
|
||||
|
||||
Once the Cacti language translation file has been properly on boarded, you can use either
|
||||
Once the Cacti language translation file has been properly onboarded, you can use either
|
||||
PO Edit or the Weblate sites to make edits to the file. We find that if you are translating
|
||||
a new language, you should first let the Cacti Group do an automatic translation, of if you
|
||||
want to contribute to PO Edit Professionals development, you can pay for a license to
|
||||
a new language, you should first let The Cacti Group do an automatic translation, or, if you
|
||||
want to contribute to PO Edit Professional's development, you can pay for a license to
|
||||
facilitate a first pass language translation.
|
||||
|
||||
One of the key benefits of using Weblate, is that if you have created additional language
|
||||
|
||||
+10
-10
@@ -6,11 +6,11 @@
|
||||
|
||||
### Background
|
||||
|
||||
PHP has matured over the years, and in the year 2020, it is now possible to
|
||||
create reliable services leveraging it. Cacti for years used crontab as a
|
||||
launcher primarily due to the issues we had experienced in the early days of
|
||||
PHP 4.2 with constant memory leaks with the core PHP and the various PHP
|
||||
modules that were in use. However, times are changing.
|
||||
PHP has matured over the years, and it is now possible to create reliable
|
||||
services leveraging it. Cacti for years used crontab as a launcher primarily
|
||||
due to the issues we had experienced in the early days of PHP 4.2 with constant
|
||||
memory leaks with the core PHP and the various PHP modules that were in use.
|
||||
However, times are changing.
|
||||
|
||||
With the advent of technologies such as MariaDB Galera, keepalived, haproxy,
|
||||
and database session management for fault tolerant web server and
|
||||
@@ -19,8 +19,8 @@ RRDfile and web site content fault tolerance, the idea of a fully
|
||||
distributed highly available Cacti is now possible.
|
||||
|
||||
Add to that, the concept of Remote Data Collectors for network resiliency,
|
||||
and Cacti has fully entering into what I refer to as 'enterprise' class in
|
||||
it's stability and reliability.
|
||||
and Cacti has fully entered into what I refer to as 'enterprise' class in
|
||||
its stability and reliability.
|
||||
|
||||
However, to more easily support keepalived, we need to move away from the
|
||||
crontab based setup Cacti has used for nearly 20 years. Long live Crontab!
|
||||
@@ -35,10 +35,10 @@ There are only a few steps. For this service here they are.
|
||||
2. Modify the file to point to the real path of the cactid.php file located
|
||||
in the cacti base directory by default
|
||||
3. Modify the file to use the user and group that your Cacti is installed
|
||||
with, on RedHat variants, it usually: apache:apache, for Debian variants,
|
||||
its generally www-run:www
|
||||
with, on RedHat variants, it is usually: apache:apache, for Debian variants,
|
||||
it is generally www-data:www-data
|
||||
4. Create the file /etc/sysconfig/cactid, even though it's not used today,
|
||||
we'll keep it there for as a future path to over write certain settings
|
||||
we'll keep it there as a future path to overwrite certain settings
|
||||
5. Install the service using 'systemctl enable cactid'
|
||||
6. Reload systemd using 'systemctl daemon-reload'
|
||||
7. Comment out the cacti crontab file or simply remove the /etc/cron.d/cacti
|
||||
|
||||
Reference in new issue
Block a user