Hello -- I upgraded (scaled up) to 24 processors and 196 Gigs of RAM on a Digital Ocean box to handle what appears to be a significant increase in traffic (wp site) when we publish an article to Facebook.
When I look at htop during the surge in traffic (after the upgrade of ram) the RAM shown is less than 5 Gigs and yet all 24 processors are swamped. "Database Unreachable" errors are shown often when attempting to get to the site.
When the traffic is low, i did the 'free' command and got what I think is a ridiculously low value for used memory -
total used free shared buff/cache available
Mem: 188Gi 2.9Gi 91Gi 389Mi 94Gi 183Gi
Swap: 0B 0B 0B
Whereas one of my other WO WP sites with eight processors and 60 gigs of RAM indicates what I typically see with very low traffic
total used free shared buff/cache available
Mem: 62Gi 15Gi 1.1Gi 550Mi 45Gi 45Gi
Swap: 0B 0B 0B
I read that I should update the MySQL stack after a memory increased, but my boxes are installed with the root user as the assigned user. That command does not work with --allow-root attached.
This site is very large in terms of data (in case I have to resort to backups) -- so i want to be sure that I can safely do a Stack Upgrade --mysql and also how can I do it as root? Can you provide the required adjustments? that I might do by hand?
I adjusted values in the mariadb.conf.d folder based on mysqltuner recommendations. But I think you also can do those automatically with your stack upgrade code -- Any help would be appreciated. The site swamps out now after publishing to facebook, every time.
I am clearly not seeing proper memory info on HTOP or TOP under load.
Thank you SL