difftaichi VS community

A bit unrelated but what would you call the learning algorithm on this example?. Its like a neural network controller but im not sure. I want to read up on it, seems pretty cool

DiffTaichi, which optimizes robots' movements, is an early-stage trial of Taichi's AutoDiff feature:

Advanced features of Taichi include spatially sparse computing and differentiable programming [examples].

Pros are instant HA and Migration. Cons are huge bandwidth hits. With your 4x1gbe you would be maxed out on replicating those 25 VMs. You wouldn't have anything for users. I have a test lab with 4 nodes, 22cpu 100gbram and 30tb space, using low end stuff, 12hdds. Proxmox, ceph dashboard, (the native ceph dashboard you can turn on), and a netdata.cloud account. So I watch it like a hawk and like to load test.

Docker-compose, not k8s. Set up a script to update the OS, pull all your containers and reboot after hours once a week or once a day. Make sure the script specifies non interactive. Set up alerting for low disk space, see https://netdata.cloud or use your own tool

There can be some issues if you mix and match elastic versions, wazuh versions, logstash versions. But the documentation guides you very well with matrix of what is and is not compatible. You will want a beefy VM to run it in, I started smaller than I should of, and after running a while it kind of puked on itself, certain things would randomly stop working. After giving it 32GB RAM, plenty of disk 4TB, and 8 vcpu, it now runs like a top. of course you will need to run, test and tune all the config files for Ossec, Wazuh, Logstash etc. The big one being JVM heap memory settings, larger these can be, the better performance. Also if you can, run it on SSD disk, helps alot being there is lots of I/O, log ingestion, processing etc. One free Linux monitoring platform I use, which I highly recommend for all Linux servers is netdata.cloud. Awesome, awesome, awesome! It actually helps tune and get real time health of the Wazuh server, and can point out pain points with hardware being undersized for example, too much swap utilization, disk I/O etc. Regardless, check that out regardless.

$ brew info netdata netdata: stable 1.29.3 (bottled) Diagnose infrastructure problems with metrics, visualizations & alarms https://netdata.cloud/ Not installed From: https://github.com/Homebrew/homebrew-core/blob/HEAD/Formula/netdata.rb License: GPL-3.0-or-later ==> Dependencies Build: autoconf ✘, automake ✘, pkg-config ✔ Required: json-c ✘, libuv ✘, lz4 ✘, [email protected] ✔ ==> Caveats To start netdata: brew services start netdata Or, if you don't want/need a background service you can just run: /usr/local/opt/netdata/sbin/netdata -D

Try netdata.cloud

What I know is that each node's data is still primarily stored on the node itself, and I've figured that the Registry used by Netdata cloud stores only URLs and randomly generated UUIDs. So my question is, will any other data be stored outside of my nodes? Does Netdata Cloud have access to my servers 24/7 or only when I got a browser tab with Netdata cloud open? Is there more information on security and data collection (besides GDPR and personal data collection) available on your site, netdata.cloud?

StatsD in Netdata

hey, sorry about taking so long to reply, and thanks for opening an issue on Gitlab. I just released v1.0.0 of https://xsrv.readthedocs.io/en/latest/ and will try to send a PR to https://github.com/netdata/community/tree/main/netdata-agent-deployment explaining how to bootsrap netdata from the ansible monitoring role in the coming weeks.