The defautl vm.swappiness value for Proxmox 5.2.x is 60
Run this command in Shell in order to reduce swappiness to an absolute necessity when the system is out of RAM.
sysctl -w vm.swappiness=0
Other performance tips
Proxmox OpenVZ SWAP and Performance
If you’re looking for a way to boost your cloud instance performance this swapiness tip might just be just the tip you’ve been looking for.
You can find in system monitor that Ubuntu is using swap space although you have plenty of free physical RAM.
The value of swappiness is between 0 ~ 100. Lower value means Linux will use swap space less whereas higher value causes Linux to use swap space more often. The default value on Ubuntu is 60 which means when your computer uses up 40% of physical RAM, Linux kernel begins swapping.
You can use the following command to check the current value.
It’s recommended to set a low value for swappiness so that Linux kernel will use as much physical RAM as possible. For example, to set swappiness to 10:
sudo sysctl vm.swappiness=10
Do not add space before and after the equal sign. This command tells Linux kernel to use swap space when 90% or more of physical RAM is used up. This setting isn’t persistent.
To make it persistent across reboots, you need to edit a file.
sudo nano /etc/sysctl.d/99-sysctl.conf
At the end of this file, append the following line.
Are your videos downloading slowing? Webpages not loading?
If you have already cleared your browser cache and have restarted your computer you might want to try this tip.
- Unplug all your modem, router. Wait 1 minute.
- Replug modem, router.
- Plug into good quality power bar for best performance.
sudo ~/Library/Google/GoogleSoftwareUpdate/GoogleSoftwareUpdate.bundle/Contents/Resources/GoogleSoftwareUpdateAgent.app/Contents/Resources/ksinstall --nuke
sudo ~/Library/Google/GoogleSoftwareUpdate/GoogleSoftwareUpdate.bundle/Contents/Resources/GoogleSoftwareUpdateAgent.app/Contents/Resources/ksinstall --uninstall
Remove anything you don’t absolutely need to run at startup or login.
Is it true that MySQL Galera is as slow as the slowest node?
Yes it is. Galera relies on group communication between nodes. For any matters, it will wait for all nodes to return the status of certification test before proceed with committing or rollbacking. At this phase, an overloaded node will surely facing a hard time to reply within the time manner, delaying the rest of the cluster to wait for it. Therefore it is recommended to have uniform servers across the cluster. Also, transaction latency is no shorter than the RTT (round trip time) to the slowest node.
You might have too many httpd processes running on your system during development.
Screenshot of httpd processes haven taken over the machine.
Objective: Reduce the amount of spare httpd servers that apache will maintain in memory.
Purpose: Since this is for local development purposes it’s not likely that the number of users will exceed 256 clients.
These are the internal default configuration for Apache sets its configuration. This section will not exist in your httpd.conf file unless you already created it prior.
Insert/replace the following in httpd.conf – File->Edit Template->Apache->httpd.conf
This will reduce the amount of spare httpd processes in memory and reduce the amounts of threads used by the processor. This will improve your overall system performance, reduce energy and battery usage as well.