Varnish 4.1.4 - MAIN.n_gunzip counter increasing for every MAIN.backend_req

[Vlad Rusu]

Thanks a lot Dridi!

That makes sense :)

To answer your 2nd question, I would have liked to see this documented as follows:

http://www.varnish-cache.org/docs/4.1/reference/varnish-counters.html <http://www.varnish-cache.org/docs/4.1/reference/varnish-counters.html> - in the n_gunzip section (i.e. includes the.. )

https://www.varnish-cache.org/docs/4.1/reference/vsl.html <https://www.varnish-cache.org/docs/4.1/reference/vsl.html> - in the Gzip tag section

https://varnish-cache.org/docs/4.1/phk/gzip.html <https://varnish-cache.org/docs/4.1/phk/gzip.html> - and here.. 


Thanks for your hard work on this cool product guys!

--
Vlad Rusu
skypeid: rusu.h.vlad | cell: +40758066019

Lola Tech | lola.tech <https://lola.tech/>
> On 06 Jan 2017, at 15:07, Dridi Boukelmoune <dridi at varni.sh> wrote:
> 
> On Fri, Jan 6, 2017 at 10:28 AM, Vlad Rusu <vlad.rusu at lola.tech> wrote:
>> Hi guys,
>> 
>> https://www.varnish-cache.org/docs/4.1/reference/vsl.html
>> 
>> Gzip tag, in the format section, shows a possible flag of ‘u’ representing a
>> Gunzip-test
>> 
>> In my varnishstat output I see how MAIN.n_gunzip equals (roughly)
>> MAIN.backend_req
>> 
>> 
>> Looking at the Gzip tag, I see the following for every backend request:
>> 
>> Gzip           u F - 70 55 80 80 489
>> 
>> 
>> Can someone help me understand what a Gunzip-test means / does? It sure gets
>> counted against MAIN.n_gunzip.
> 
> Hello Vlad,
> 
> The gunzip test is performed when an already-gzipped backend response
> is inserted in the cache. It is inflated in a black-hole, only to make sure
> that the object being fetched is proper gzip.
> 
>> Am not attaching the VCL or a full varnishstat as I don’t think it makes
>> sense. Each backend sends gzipped data back to Varnish and each client
>> requests for gzipped data from Varnish..
> 
> For each gzipped backend response, you will see this behavior (u flag
> in logs, and counter increment). There is no separate counter for test
> gunzip vs regular gunzip, although some of us would like to make the
> distinction.
> 
> As an end-user, how and where would this be best documented in your
> opinion?
> 
> Best Regards,
> Dridi

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.varnish-cache.org/lists/pipermail/varnish-misc/attachments/20170106/712c9628/attachment.html>