Hey Eray,
Indeed all of these are good and sysadmins should be able to handle them
but.. in specific cases it's not easy.
The cases I know about are:
- SAT links (slow or costly)
- Sensitive acl\security systems
- Very low quality distance wireless links
In the case of ACLs system bypass or bridging might not be an option if
the stakes are high(This is where I asked myself couple times why IT
managers doesn't like to pay for industrial guarantees).
I am happy with squid for a very long time and I couldn't understand why
a friend of mine wasn't happy about it. Only when I was with him and he
showed me what happens when he tries to install and run squid I
understood the Blocking issue. Eventually for him a special customized
proxy was the answer.
Thanks,
Eliezer
On 01/03/2016 12:55, Eray Aslan wrote:
False dichotomy. There is always something you can do. Re-route the
traffic, throw the bypass switch, bridge the interfaces, don't use the
cache, downgrade, take preventive measures uptream in the flow...
i.e. let the sysadmin/system architect handle the emergencies. The case
above is not different from squid box(es) going offline for whatever
reason. Worst case: Live through the outage, learn from it and
hopefully design your systems accordingly in the future.
-- Eray
_______________________________________________
squid-users mailing list
squid-users@xxxxxxxxxxxxxxxxxxxxx
http://lists.squid-cache.org/listinfo/squid-users
