inconsistent output from debug.backend / debug.health

[Poul-Henning Kamp]

In message <CA+3979=17BERgc+EitGcnLZMt+sCSpa1XX+K28ZWeqMJZZr-xQ at mail.gmail.com>
, Nathan Mehl writes:

>Context: we have varnish fronting for a pool of servers that
>(semi-)automatically scale based on load criteria. A puppet job updates a
>vcl file defining all of the directors any time a server is added or
>removed from the pool, and then reloads the VCL.

We have recently added an "administrative health" setting to Varnish
and I think you should use that instead to selectively enable and
disable backends.

>debug.backend:
>
>0x7f83618dfc00 cms_director[0](10.251.39.31,(null),:8080) 3 0
>0x7f51686f8840 cms_director[1](10.62.87.78,(null),:8080) 2 2
>0x7f516848f480 cms_director[0](10.62.87.78,(null),:8080) 5 1

We try to recognize when a new VCL has the _exact_ same backend
definition as another loaded VCL, so that we can reuse that
instance of the backend (and any cached open connections etc)

For that to work, the backend definition must be _exactly_ the
same, token for token.


-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.