[squid-users] Does reload_into_ims not effective for HTTPS?

[Amos Jeffries]

On 18/03/2016 11:49 p.m., Yuri Voinov wrote:
> Hi gents,
> 
> I see strange behaviour for many URL's.
> 
> Im my setups reload_into_ims is on globally.
> 
> This time I see:
> 
> - when I pressed Ctrl+F5 in browsers, for HTTP URLs
> 
> TCP_REFRESH_UNMODIFIED/200 6447 GET
> http://icdn.lenta.ru/images/2016/03/17/09/20160317091221731/tabloid_8a08b3a372ff4499c0d95723ad4dc382.jpg
> - ORIGINAL_DST/81.19.76.10 image/jpeg
> 
> occurs.
> 
> - when I goes under HTTPS for the same URLs, i see TCP_MISS:
> 
> TCP_MISS/200 6447 GET
> https://icdn.lenta.ru/images/2016/03/17/09/20160317091221731/tabloid_8a08b3a372ff4499c0d95723ad4dc382.jpg
> - ORIGINAL_DST/81.19.76.10 image/jpeg
> 
> When I've test HTTPS URL with redbot, I've seen no problem with caching
> content:
> 
> http://i.imgur.com/9dm7VWV.png
> 
> As you can see, this site is news, with no private headers and no
> authentication.
> 
> What's up? Is it bug or documented behaviour? I see absolutely no
> problem with caching either http or https versions of URL.
> 

What headers are on the two requests?

Did the cache contain the https:// URL object prior to the refresh?

Amos