On 07/11/2014 05:27 AM, Tsantilas Christos wrote: > The PageSpeed example fits better to a post-cache RESPMOD feature.
I do not think so. Post-cache RESPMOD does not allow Squid to cache the adapted variants. Please let me know if I missed how post-cache RESPMOD can do that. The key here is that PageSpeed and similar services want to create (and cache) many adapted responses out of a single virgin response. Neither HTTP itself nor the Squid architecture support that well. Post-cache REQMOD allows basic PageSpeed support (the first request for "small" adapted content gets "large" virgin content, but the second request for small content fetches it from the PageSpeed cache, storing it in Squid cache). To optimize PageSpeed support further (so that the first request can get small content), we will need to add another generally useful feature, but I would rather not bring it into this discussion (there will be a separate RFC if we get that far). The alternative is to create a completely new interface (not a true vectoring point) that allows an adaptation service to push multiple adapted responses into the Squid cache _and_ tell Squid which of those responses to use for the current request. While I have considered proposing that, I still think we would be better off supporting "standard" and "well understood" building blocks (such as standard adaptation vectoring points) rather than such highly-specialized interfaces. Please let me know if you disagree. > Is > the post-cacge REQMOD just a first step to support all post-cache > vectoring points? You can certainly view it that way, but I do not propose or promise adding post-cache RESPMOD :-). Thank you, Alex. > On 07/11/2014 01:15 AM, Alex Rousskov wrote: >> Hello, >> >> I propose adding support for a third adaptation vectoring point: >> post-cache REQMOD. Services at this new point receive cache miss >> requests and may adapt them as usual. If a service satisfies the >> request, the service response may get cached by Squid. As you know, >> Squid currently support pre-cache REQMOD and pre-cache RESPMOD. >> >> >> We have received many requests for post-cache adaptation support >> throughput the years, and I personally resisted the temptation of adding >> another layer of complexity (albeit an optional one) because it is a lot >> of work and because many use cases could be addressed without post-cache >> adaptation support. >> >> The last straw (and the motivation for this RFC) was PageSpeed[1] >> integration. With PageSpeed, one can generate various variants of >> "optimized" content. For example, mobile users may receive smaller >> images. Apache and Nginx support PageSpeed modules. It is possible to >> integrate Squid with PageSpeed (and similar services) today, but it is >> not possible for Squid to _cache_ those generated variants unless one is >> willing to pay for another round trip to the origin server to get >> exactly the same unoptimized content. >> >> The only way to support Squid caching of PageSpeed variants without >> repeated round trips to the origin server is using two Squids. The >> parent Squid would cache origin server responses while the child Squid >> would adapt parent's responses and cache adapted content. Needless to >> say, running two Squids (each with its own cache) instead of one adds >> significant performance/administrative overheads and complexity. >> >> >> As far as internals are concerned, I am currently thinking of launching >> adaptation job for this vectoring point from FwdState::Start(). This >> way, its impact on the rest of Squid would be minimal and some adapters >> might even affect FwdState routing decisions. The initial code name for >> the new class is MissReqFilter, but that may change. >> >> >> >> The other candidate location for plugging in the new vectoring point is >> the Server class. However, that class is already complex. It handles >> communication with the next hop (with child classes doing >> protocol-specific work and confusing things further) as well as >> pre-cache RESPMOD vectoring point with caching initiation on top of >> that. The Server code already has trouble distinguishing various content >> streams it has to juggle. I am worried that adding another vectoring >> point there would make that complexity significantly worse. >> >> It is possible that we would be able to refactor/encapsulate some of the >> code so that it can be reused in both the existing Server and the new >> MissReqFilter classes. I will look out for such opportunities while >> trying to keep the overall complexity in check. >> >> >> Any objections to adding post-cache REQMOD or better implementation >> ideas? >> >> >> Thank you, >> >> Alex. >> [1] https://developers.google.com/speed/pagespeed/ >>