Re: Diazo overhead?
Martin Aspeli <[email protected]>
| Newsgroups | gmane.comp.web.zope.plone.user |
|---|---|
| Message-ID | <CAPN0AAQsB1w1f1yLQBcDDYJAztrrtXSQrPnKwzoX5jafwJUPBQ@mail.gmail.com> |
On 5 November 2014 19:36, Sean Upton <[email protected]> wrote: > Diazo newbie question: > > What's the typical overhead for Diazo to run a transform on a very > typical page in Plone? I have not been able to determine this, and > have not yet needed to invest the effort to profile it against some > non-trivial theme. > Pretty low. It's pretty well optimised (unless you have very complex or very open-ended XPath queries). Probably much less than whatever else you're doing on the page. But you always need to benchmark your particular use case and try not to optimise prematurely. > In trying to understand how plone.app.theming (and > plone.transformchain) works, AFIACT it looks like there is no > caching/memoization of a transform result, even if the response body > prior to transform is byte-for-byte identical to a previously > transformed request in the same instance (not in development mode; > only the transforms themselves are cached, not the result). > No, because there's not been a need to optimise that bit yet. Generally, if this were the case, you'd cache the page anyway and/or pass a 304 response so the request would never even hit Plone. Martin ------------------------------------------------------------------------------ _______________________________________________ Plone-Users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/plone-users