Questions[cache](server)

We just migrated to Amazon AWS. We currently have an EC2 instance that's working well. It's running Nginx in front and Apache in the back-end. That's running well also. All sites are launched properly and includes the Cache-Control header for files that are served from the EC2.

The problem is with ALL static files we placed in Amazon S3 that's being accessed through CloudFront CDN. We can access the files fine (and no issue with CORS), but apparently CloudFront doesn't serve files with Cache-Control header. We want to leverage on browser caching.

The way I see it, the EC2 instance doesn't play a role here as the static files are being served directly by S3+CloudFront, the request does not go to the Web Server in EC2.

I'm at a complete lost.

Question: 1) How do I set the Cache-Control in this case? 2) Is it possible to set the Cache-Control? From S3 or CloudFront?

Note: I've hit a few pages in Google where you can set the Header in S3 for individual objects. That's really not a productive way to do it specially since in my case we are talking of several objects.

Thanks!

I have a HAProxy / stunnel server that handles SSL for our sites on AWS. During testing, I created a self-signed cert on this server and hit it from my desktop using Chrome to test that stunnel was working correctly.

Now I have installed the legitimate cert on that server. When I hit the site from my machine in Chrome it throws the following error:

Error 113 (net::ERR_SSL_VERSION_OR_CIPHER_MISMATCH): Unknown error.

My guess is that Chrome cached the key for the self-signed cert and it doesn’t match that of the legitimate cert. This site works in all other browsers on my machine so it’s just a Chrome problem.

One interesting note: When hitting the page from a incognito session (Ctrl+Shift+N), it works correctly. So it is clearly some sort of cache thing.

I did all the things I could think of (dumped my cache, deleted certs from the Personal and Other People page in the Manage Certificates dialog, Ctrl+F5, etc.).

My machine is Windows 7 x64. Chrome version: 12.0.742.91.

On the Google Chrome Help Form, there is a description of what sounds like the same issue; however, no resolution is found.

UPDATE: It seems to have “fixed itself” today. I hate problems like this. I still don’t know what caused it or how it resolved itself. Presumably the cached cert expired or something, but I am still interested to know where this information is stored and how to verify it.

We are using Nginx to serve static files on a development platform. As it is a development platform, we'd like to disable caching so that each change is propagated to the server. The configuration of the VHost is quite simple:

server {
  server_name  static.server.local;
  root /var/www/static;

  ## Default location
  location / {
    access_log        off;
    expires           0;
    add_header        Cache-Control private;
  } 
}

When we access an HTML file (http://static.server.local/test.html), we have no issue: the server returns a code 304 Not Modified as long as the file is not changed, and a 200 OK response with the modified file when the file is changed.
However, it seems to behave differently with a Javascript or a CSS file. Once the file is changed, we get a 200 OK response as expected, but with the old text.
Is there an internal cache mechanism in Nginx that could explain this behaviour? Or some configuration that we should add?

As a side note, here is the header returned by Nginx when the file has been modified (it seems correct):

Accept-Ranges:bytes
Cache-Control:max-age=0
private
Connection:keep-alive
Content-Length:309
Content-Type:text/css
Date:Fri, 13 May 2011 14:13:13 GMT
Expires:Fri, 13 May 2011 14:13:13 GMT
Last-Modified:Fri, 13 May 2011 14:13:05 GMT
Server:nginx/0.8.54

Edit
After trying different settings with the expires directive and Cache-Controlheader, I made some further investigations. In fact, the server is installed on a VirtualBox guest Ubuntu, and data are read from a shared folder that is on the Mac OSX host.
If the file is edited from an IDE (NetBeans) on the host, it seems that changes do not appear whereas if I edit it directly on the guest (using VIM), it is refreshed.
The strange thing is it does not behave similarly with HTML files.
Quite puzzling.

Edit 2 (ANSWER)
Indeed, the origin of the issue was more on the VirtualBox side. Or rather a conflict between VirtualBox and the "sendfile" option of the server.
This link VirtualBox Hates Sendfile gave me the solution: switch the sendfile flag in the server configuration to off:

sendfile  off;

Hope this could also help other person using VirtualBox for development. :)
There are some additional information on the VirtualBox forum.

I keep hearing about some PHP (opcode) caches like - APC, XCache, Memcache, eAccelerator, etc.

But I couldn't ever figure out how to go about choosing one. Apart from performance benefit, which a caching system is supposed to deliver, which other factors should be a point of concern.

Like why you will say X cache system is better than Y? I am less worried about relative performance gain. Small differences between any two systems matter less.

If a generic answer to my question is not possible, here are few pointers. I use dedicated VPS with Mediatemple (with root access). RAM is 512 MB (physical) + 400MB (swap) I am concerned about WordPress and its cousins WordPress-MU and BuddyPress. 90% of our codes/sites fall into WordPress family.

Thanks in advance for some help.

I have a rather old server that has 4GB of RAM and it is pretty much serving the same files all day, but it is doing so from the hard drive while 3GBs of RAM are "free".

Anyone who has ever tried running a ram-drive can witness that It's awesome in terms of speed. The memory usage of this system is usually never higher than 1GB/4GB so I want to know if there is a way to use that extra memory for something good.

• Is it possible to tell the filesystem to always serve certain files out of RAM?
• Are there any other methods I can use to improve file reading capabilities by use of RAM?

More specifically, I am not looking for a 'hack' here. I want file system calls to serve the files from RAM without needing to create a ram-drive and copy the files there manually. Or at least a script that does this for me.

Possible applications here are:

• Web servers with static files that get read alot
• Application servers with large libraries
• Desktop computers with too much RAM

Any ideas?

Edit:

• Found this very informative: The Linux Page Cache and pdflush
• As Zan pointed out, the memory isn't actually free. What I mean is that it's not being used by applications and I want to control what should be cached in memory.