Richard Arends

> Sorry and how do I switch Squid to SMP mode to test this? That's very simple, just put `workers 2' in squid.conf. See also http://www.squid-cache.org/Versions/v4/cfgman/workers.html . I have 4...

I adopted to the changes in Home Assistant and placed all the teslamate related configurations in one file, so that I could include it as a package: In configuration.yaml: ```lang=yaml...

I tested with the complete config and then I still see the memory leak. Today I will run with only the systemd part to see if there is a change.

> > > I'm investigating similar issue from few days, and observed that for fluentd `1.11.5` using `null` output with file buffer also constantly takes more memory in opposition to...

> > > It seems related to #3342. > > FYI: Recently, fluent-plugin-systemd 1.0.5 was released. > > It fixes "Plug a memory leaks on every reload" > [fluent-plugin-systemd/fluent-plugin-systemd#91](https://github.com/fluent-plugin-systemd/fluent-plugin-systemd/pull/91) >...

Thanks to @sumo-drosiek, I can confirm that no memory leak occurs when I do not use a buffer section in the output. I rerun everything using a new config with...

@ashie I just wanted to add the results of my test run with buffering configured in an output where I see that without buffering there is no memory leak and...

So to be complete: No memory leak: ```` @type null ```` Memory leak: ```` @type null @type "file" path "/var/log/fluentd-buffers/output.buffer" compress gzip flush_interval 5s flush_thread_count 8 chunk_limit_size 1m total_limit_size 128m...

The following graph displays the memory over almost 7 days. This is my regular config where only the buffer sections in the Elasticsearch outputs is disabled. I was already sure...

The known_devices.yaml file is deprecated, but still working. So I still have a small known_devices.yaml file with only the tesla_location entry in it. ```lang=shell $ cat known_devices.yaml tesla_location: name: tesla_location...