Vary with Care

Reflective Perspective - Chris Alcock » The Morning Brew #371

Reflective Perspective - Chris Alcock » The Morning Brew #371 — Thu, 18 Jun 2009 10:23:05 GMT

PingBack from http://blog.cwa.me.uk/2009/06/18/the-morning-brew-371/

Findings of June 22 - Jack is Here

Findings of June 22 - Jack is Here — Mon, 22 Jun 2009 10:57:42 GMT

PingBack from http://jack-fx.com/interesting-links/4/findings-of-june-22

re: Vary with Care

Steve Clay — Tue, 21 Jul 2009 21:17:37 GMT

In a response, Vary indicates content negotiation is occurring, so "Vary: Accept-Encoding" should be sent even when the non-encoded version is sent.

If you don't, proxies may end up caching only the non-encoded version. <a href="http://wiki.squid-cache.org/KnowledgeBase/VaryNotCaching">Squid does this</a>.

What I haven't figured out is what to do about IE6 without the "SV1" in the UA. You can't reliably send gzipped JS to it, but if you leave off Vary and the request passes through a proxy, you've eliminated the ability of the cache to store the gzipped version.

re: Vary with Care

EricLaw [MSFT] — Tue, 21 Jul 2009 21:30:53 GMT

@Steve Clay: From a protocol-perspective, of course you are correct. The point of this post is to explicitly state that using Vary: Accept-Encoding on uncompressed responses hurts performance in the most popular user-agents.

Squid could work around their limitation in the most common case (e.g. don't blindly delete compressed, cached variants that are still fresh).

An IE6 user who is current on patches should be able to properly decompress any response -- XPSP2 (identified by the SV1) shouldn't be required. In contrast, an IE6 user who is not current on patches has probably already been pwned by the bad guys.

re: Vary with Care

Benjamin — Thu, 30 Jul 2009 16:26:15 GMT

You wrote : "unless the Vary header contains only the token User-Agent or Accept-Encoding."

And if the Vary header is:

"Vary: Accept-Encoding,User-agent"

does IE6 treat the response as uncacheable?

Thanks in advance

re: Vary with Care

Mark Nottingham — Mon, 19 Oct 2009 01:50:26 GMT

Eric,

I'm starting to think that IE6's behaviour is preferable to IE7's.

Since it (presumably) controls the UA and AE headers, and doesn't allow their modification (using XHR, etc.), IE6 can safely strip Vary: UA and AE, and treat those responses as fresh. IE7, OTOH, forces revalidation on these responses.

Why not combine the behaviours and force revalidation on anything with a Vary that's *not* AE or UA? That seems like the most performance and safest thing you can do without remembering request headers...

Also, I don't think it's as easy as you say for Squid to correct their behavour; there's no simple way for them to differentiate between a server that doesn't set correct headers and one that has stopped supporting compression.

Cheers,

re: Vary with Care

EricLaw [MSFT] — Mon, 19 Oct 2009 15:05:50 GMT

@Mark: IIRC, IE7 & IE8 also ignore the Vary: User-Agent directive.

It's also important to recall that IE6 doesn't simply revalidate Vary'ing responses; it doesn't cache them at all.

re: Vary with Care

Mark Nottingham — Tue, 20 Oct 2009 08:36:21 GMT

Hi Eric,

Right, but IE6 will cache something that varies on AE or UA, right? Do 7 and 8 also ignore Vary: Accept-Encoding? If so, this isn't a big problem in the common(est) case...

Cheers,