# Is there such a thing as server-side caching? **URL:** \ **Category:** Factual Questions\ **Created:** [January 31, 2012, 5:28am UTC](https://boards.straightdope.com/t/is-there-such-a-thing-as-server-side-caching/611280 "2012-01-31T05:28:18Z")\ **Posts on this page:** 13\ **Page:** 1
**Author:** ![pulykamell](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/pulykamell/32/3166_2.png) [@pulykamell](https://boards.straightdope.com/u/pulykamell)\ **Post date:** [January 31, 2012, 5:28am UTC](https://boards.straightdope.com/t/is-there-such-a-thing-as-server-side-caching/611280/1 "2012-01-31T05:28:18Z")
I assume there is, because I otherwise have no explanation. This is the second time the following has happened to me: FatCow is my hosting provider. I uploaded a webpage that consisted of thumbnails with links to larger images. I noticed that one of the pages I uploaded had the wrong photo on it, so I went back in to upload the correct photo in both the thumbnail and full photo directories. When I access the updated page, the thumbnail shows as the new photo, but it still links to the old content. OK, says I. Clear the browser cache. Same issue. So I start up another browser that never accessed that web page. Same issue. So I go to another computer. Same thing. So I try the iPhone via 3G (i.e., through a different ISP than my cable internet). Yep, same thing. I rename the file to something like “photo3b” instead of “photo3.” Hey, the updated file loads up. So the right file is on the server, it’s just that whenever I try to access it from wherever, the old one is fed to me. Yes, I just opened up a support ticket to ask what’s going on. In the meantime, I assume some of you smart people know what’s going. I’m assuming it must be server-side caching, because what else could it be? Is this usual? Is there a way to force it not to cache? ---
**Author:** ![Askance](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/askance/32/8281_2.png) [@Askance](https://boards.straightdope.com/u/Askance)\ **Post date:** [January 31, 2012, 5:34am UTC](https://boards.straightdope.com/t/is-there-such-a-thing-as-server-side-caching/611280/2 "2012-01-31T05:34:21Z")
Of course there is. If each write to a multi-user server was forced to commit to physical disk at write-time the server’s performance would be atrocious. That said, if you are getting this issue over a period of more than a day there’s more going on than just failure to write to disk from cache. But there’s nothing you can do about it from your end. ---
**Author:** ![TriPolar](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/tripolar/32/3008_2.png) [@TriPolar](https://boards.straightdope.com/u/TriPolar)\ **Post date:** [January 31, 2012, 5:47am UTC](https://boards.straightdope.com/t/is-there-such-a-thing-as-server-side-caching/611280/3 "2012-01-31T05:47:42Z")
Sounds like server side caching. They may have a problem accessing the actual file, and you’ll keep getting the cache version until it becomes available, if ever. ---
**Author:** ![Rhythmdvl](https://avatars.discourse-cdn.com/v4/letter/r/85f322/32.png) [@Rhythmdvl](https://boards.straightdope.com/u/Rhythmdvl)\ **Post date:** [January 31, 2012, 5:54am UTC](https://boards.straightdope.com/t/is-there-such-a-thing-as-server-side-caching/611280/4 "2012-01-31T05:54:19Z")
I had a similar question a year or so ago. You may want to take a look through: [Do ISPs cache pages—and is there anything I can do about it?](http://boards.straightdope.com/sdmb/showthread.php?t=585914) ---
**Author:** ![pulykamell](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/pulykamell/32/3166_2.png) [@pulykamell](https://boards.straightdope.com/u/pulykamell)\ **Post date:** [January 31, 2012, 6:00am UTC](https://boards.straightdope.com/t/is-there-such-a-thing-as-server-side-caching/611280/5 "2012-01-31T06:00:08Z")
> [@Rhythmdvl](#): > > I had a similar question a year or so ago. You may want to take a look through: > > [Do ISPs cache pages—and is there anything I can do about it?](http://boards.straightdope.com/sdmb/showthread.php?t=585914) Interesting. This was my first thought (well, second, after browser caching), but that’s why I went to my iPhone to rule that possibility out. This is interesting to me, as my last hosting service did not seem to cache content. When I made a change, it went live right away. ---
**Author:** ![pulykamell](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/pulykamell/32/3166_2.png) [@pulykamell](https://boards.straightdope.com/u/pulykamell)\ **Post date:** [January 31, 2012, 6:02am UTC](https://boards.straightdope.com/t/is-there-such-a-thing-as-server-side-caching/611280/6 "2012-01-31T06:02:05Z")
> [@Askance](#): > > Of course there is. If each write to a multi-user server was forced to commit to physical disk at write-time the server’s performance would be atrocious. > > That said, if you are getting this issue over a period of more than a day there’s more going on than just failure to write to disk from cache. But there’s nothing you can do about it from your end. So, when somebody makes a change to their website content, it could take up to a day to propagate? I’m sure I must be misunderstanding something. ---
**Author:** ![pulykamell](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/pulykamell/32/3166_2.png) [@pulykamell](https://boards.straightdope.com/u/pulykamell)\ **Post date:** [January 31, 2012, 6:04am UTC](https://boards.straightdope.com/t/is-there-such-a-thing-as-server-side-caching/611280/7 "2012-01-31T06:04:57Z")
Ah, it finally flushed its content. The new version is now showing up. So looks like about an hour or so from when I changed it to when the cache is updated. ---
**Author:** ![TriPolar](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/tripolar/32/3008_2.png) [@TriPolar](https://boards.straightdope.com/u/TriPolar)\ **Post date:** [January 31, 2012, 6:12am UTC](https://boards.straightdope.com/t/is-there-such-a-thing-as-server-side-caching/611280/8 "2012-01-31T06:12:18Z")
> [@pulykamell](#): > > So, when somebody makes a change to their website content, it could take up to a day to propagate? I’m sure I must be misunderstanding something. There isn’t any specific time. It’s just based on the load on the server. Cache updates are a low priority, and any case might have been due to a problem on the server storage. They may simply have been backing up the disk holding your file. ---
**Author:** ![pulykamell](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/pulykamell/32/3166_2.png) [@pulykamell](https://boards.straightdope.com/u/pulykamell)\ **Post date:** [January 31, 2012, 6:23am UTC](https://boards.straightdope.com/t/is-there-such-a-thing-as-server-side-caching/611280/9 "2012-01-31T06:23:44Z")
> [@TriPolar](#): > > There isn’t any specific time. It’s just based on the load on the server. Cache updates are a low priority, and any case might have been due to a problem on the server storage. They may simply have been backing up the disk holding your file. So, out of curiosity, how does it work with time-sensitive content like at news sites? ---
**Author:** ![jjimm](https://avatars.discourse-cdn.com/v4/letter/j/ba8739/32.png) [@jjimm](https://boards.straightdope.com/u/jjimm)\ **Post date:** [January 31, 2012, 6:33am UTC](https://boards.straightdope.com/t/is-there-such-a-thing-as-server-side-caching/611280/10 "2012-01-31T06:33:16Z")
> [@TriPolar](#): > > There isn’t any specific time. It’s just based on the load on the server. Not always. The site I used to run (on SharePoint) had a scheduled hourly de-cache. I’d imagine on news sites they de-cache every few minutes, or have a ‘top story’ de-cache trigger. ---
**Author:** ![pulykamell](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/pulykamell/32/3166_2.png) [@pulykamell](https://boards.straightdope.com/u/pulykamell)\ **Post date:** [January 31, 2012, 6:35am UTC](https://boards.straightdope.com/t/is-there-such-a-thing-as-server-side-caching/611280/11 "2012-01-31T06:35:42Z")
This is all very enlightening. I don’t know why I hadn’t noticed it before. Thanks. ---
**Author:** ![TriPolar](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/tripolar/32/3008_2.png) [@TriPolar](https://boards.straightdope.com/u/TriPolar)\ **Post date:** [January 31, 2012, 6:40am UTC](https://boards.straightdope.com/t/is-there-such-a-thing-as-server-side-caching/611280/12 "2012-01-31T06:40:40Z")
> [@pulykamell](#): > > So, out of curiosity, how does it work with time-sensitive content like at news sites? > [@jjimm](#): > > Not always. The site I used to run (on SharePoint) had a scheduled hourly de-cache. > > I’d imagine on news sites they de-cache every few minutes, or have a ‘top story’ de-cache trigger. Sure. Any server may have a time limit on caching. I was speaking generally. Depending on the application there may be differing priorities as well. Web pages and pictures are often on a lower priority. But even a news site may have problems that interfere with cache updates. Besides increasing performance, one of the functions of caching is to provide a second source for data when there are problems. In addition, there are mirroring systems and other techniques for data integrity so that the data accessed isn’t the most current version. ---
**Author:** ![Mangetout](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/mangetout/32/19_2.png) [@Mangetout](https://boards.straightdope.com/u/Mangetout)\ **Post date:** [January 31, 2012, 8:29am UTC](https://boards.straightdope.com/t/is-there-such-a-thing-as-server-side-caching/611280/13 "2012-01-31T08:29:00Z")
There’s another possible explanation - depending on the server architecture, the file system may be case-sensitive, so photo3 and Photo3 would not be the same thing - and renaming the files in the described manner might fix the problem without ever revealing what was causing it. Or it might be something else - I mention this only because it _kind of_ fits the phenomenon as described (there may be some obvious undescribed detail that eliminates this possibility)