Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

I have a website that runs on 4 loadbalanced servers. These are Windows servers, each with 16 GB of memory. I need to move to a shared cache, so I was thinking of Memcached. However, budgetary reasons do not allow me to add extra servers. Since my 4 webservers already have 16 GB each, I was thinking of using each webserver as a Memcached node instead of running memcache on 1 or 2 dedicated servers. So the 4 webservers will have a single, shared cache which is distributed among the same 4 servers, using Memcached.

Is this wise? What are the disadvantages? Will I run into problems?

share|improve this question
add comment

1 Answer

We've implemented something similar, but with three (two active, one passive) web servers, with the default of 64MB allocated to one instance of memcached per server. We're fairly new to it, but overall, it has been working for us.

Some things you have to think about are how much memory is going to be used by memcached. If you have plenty to spare, then you should be in good shape.

The disadvantages are conditional, for e.g., if you don't have enough memory/resources or your application's use of memcached is inefficient.

You might run into problems (-: But be ready to diagnose and troubleshoot. Join the memcached mailing list. Google "memcache.php" or "php memcache admin" for some basic monitoring tools. Also, look at some of the tools that come with memcached as such as memstat, memcached-tool, etc. If you use Nagios or something similar, there might be some options there as well.

If you have it, try this in a test server/environment first to see how set reacts to memcached.

Good luck and HTH.

share|improve this answer
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.