graph: Write the graph to the output stream

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

graph: Import the default arguments from the Python implementation

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

Make Gettext available

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

Import some colours from the Python implementation

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

graph: Fix calling rrd_graph()

There is pretty much no documenation of librrd and the API has been
changed multiple times. Initially, various arguments have been labeled
optional, but that does not seem to be true for at least the
implementation I am testing against.

So we pass all variables that we function has, and we might as well
include them in the debugging output.

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

args: Add function to add multiple pre-formatted arguments at once

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

module: Remove hack where the event loop stores the reference

This was needed when we did not have a proper registry for all modules,
but it seems that that is now the easier way.

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

graph: Pass the object to the render function

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

graph: Add a helper function to pull in all data sources from a module

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

graph: Add a render function to jump into the individual graphs

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

modules: Add function to fetch a specific module by its name

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Add function to fetch all modules

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

graph: Create the basic to render a graph

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

graphs: Fix typo in variable name

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Create a modules registry

Just like we have for graphs now

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

graphs: Take the object name as first (and only) argument

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

graph: Add a Render() method to the bus to render graphs

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

graphs: Export them over dbus

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

graphs: Add scaffolding for a loadavg graph

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

jenkins: Don't fail if the static analyzer didn't find any issues

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Initialize all graphs when the daemon starts

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Initialize all graphs when the daemon starts

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Generalize the init/exit functions

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

graphs: Add scaffolding for a graph collection

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

main: Log an error if we could not free all resources

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

graphs: Add scaffolding for a new object

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Add a basic dbus interfaces for graphs

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

queue: Allocate the correct amount of memory

This has been found by the clang static analyzer.

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

jenkins: The configure script does not have a --enable-debug switch

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

configure: Make lm-sensors an optional dependency

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

configure: Check for pkg-config before using it

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

configure: Make libatasmart an optional dependency

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

configure: Drop dependency to liboping

This won't be implemented in this daemon any more, but move to the
network daemon which is broadcasting the results over the bus.

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

Drop the Python implementation

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Export the version over dbus

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

Add a Jenkinsfile

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

po: Update langauge files

This is mainly to make "make distcheck" happy.

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Connect to dbus

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

module: Explicitely clear any RRD errors

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

Make secure_getenv available

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

modules: Add the conntrack module

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

util: Add helper function to read integers from file

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

module: Fix the heartbeat of the RRD database

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

modules: Add context switches

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

queue: Commit only once every 5 minutes

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

module: Define some default RRAs

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

module: The step size is not passed as a command line argument

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

module: Call RRD tool to create a new database

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

args: Fix typo in function name

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

module: Add the step size to the RRD configuration

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

module: Format the data sources

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

util: Add helper function to format numbers for RRD

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

modules: Rename RRD schema

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

args: Add function to dump the array

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

args: Make space in the array for the NULL termination

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

args: Add functions to fetch values

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

args: Implement pushing arguments to the array

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Create scaffolding to format argument lists

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

modules: Define the RRD schema

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

module: Create scaffolding to create the database if it does not exist

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

module: Implement batched committing of samples to the database

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

Link against rrdtool

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

queue: Store the timestamp with the sample

We can no longer store it this way because we are storing multiple
samples in a single object now.

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

queue: Store multiple samples in one queue object

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

queue: Flush on exit

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Move the queue into a separate object

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

Actually free the objects

I don't know why I keep forgetting this.

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Flush the queue on SIGHUP

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Call a function to regularly flush the queue

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Cleanup the queue on exit

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

module: Format any arguments on the stack

We seem to be keeping the allocator quite busy, which we don't need to
do too much.

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Store any submitted values

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

loadavg: Actually read the loadavg

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

module: Create a collect() function

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

modules: Create a dummy function for modules to submit data

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

modules: Call a heartbeat function every 60s

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

modules: Have each module call their own destructor

That allows us to not keep any references to the modules.

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

modules: Make them a full-blown object

This way we can store some module-specific things.

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

modules: Create a dummy loadavg module

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

logging: Always compile in debug logging

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Create some scaffolding for modules

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Terminate on SIGTERM/SIGINT

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Enable the watchdog

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Create an event loop

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

logging: Make all log levels usable

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

main: Initialize the daemon

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Create more daemon scaffolding

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Implement logging

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Link against systemd

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Parse command line options

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

ctx: Create a dummy function to set the log level

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Create a context

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

daemon: Create scaffolding

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

_collecty: Fix compiling the module with the new flags

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

configure: Configure the compiler

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

ping: Import sys/time.h for gettimeofday

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

configure: Always require systemd

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

sensors: Fix compilation with newer versions of Python

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

configure: Require at least Python 3.9

We do not use any features that are not available in an earlier version,
but I am using this for development and so didn't test anything earlier.

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

configure: Check for rrdtool Python module

Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>