Pushpendra Singh Chauhan Posted April 12, 2019 Share Posted April 12, 2019 Hi, I have a huge community having DB Size over 16Gb, daily page views over 500k. Average online users 10000-15000 on peaks and around 500K daily pageviews. Now I am expecting over 1 million online users on a particular day. This is the major event and I don't want any performance glitch. The Server detail is: CPU - 16, RAM - 64GB PHP - v7.1.23, MySQL - v5.7.24-log Please suggest what should be the mitigation strategies Link to comment Share on other sites More sharing options...
RevengeFNF Posted April 12, 2019 Share Posted April 12, 2019 Going from 15k peak users to 1 Million users in a particulary day will act almost as a ddos attack to your site. Basically you need better hardware. If they are guest users, enable the cache mechanisms in IPS so they see a cached page. Link to comment Share on other sites More sharing options...
Management Matt Posted April 12, 2019 Management Share Posted April 12, 2019 To cope with that kind of spike, you'll really need auto-scaling cloud hosting so it can increase capacity on demand. You'll want to cache as much data as you can, and consider a CDN to offload resource loading freeing up the server to focus on PHP and MySQL. Link to comment Share on other sites More sharing options...
Pushpendra Singh Chauhan Posted April 12, 2019 Author Share Posted April 12, 2019 @ RevengeFNF Thanks for the reply. All of them will be the guest users. As of now No caching is enabled. Please suggest what would be better Memcached, Redis or XCache? For this number of users could you please suggest the appropriate hardware. Thanks Link to comment Share on other sites More sharing options...
bfarber Posted April 12, 2019 Share Posted April 12, 2019 If a large majority of the users will be guests, I would set up Redis and then also ensure your configuration is set to use Redis for guest page output caching. You can increase the amount of time to cache a page (say from 30 seconds to 60 seconds) in the process, and the result will be that once a page is cached (for 60 seconds), subsequent guest requests for that page will pull from Redis and never touch your database. The PHP overhead is minimal in this case, since we fetch the cache very early on in execution. Link to comment Share on other sites More sharing options...
Pushpendra Singh Chauhan Posted April 13, 2019 Author Share Posted April 13, 2019 @ On 4/12/2019 at 3:23 PM, Matt said: To cope with that kind of spike, you'll really need auto-scaling cloud hosting so it can increase capacity on demand. You'll want to cache as much data as you can, and consider a CDN to offload resource loading freeing up the server to focus on PHP and MySQL. Thanks for your suggestion Could you suggest a good CDN ? Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.