Hi Amos,
Thanks a lot for your detailed explanation, I believe that I had big
misunderstanding of how Classes work.
Having said that, I am not sure if class 4 will be the best one for me
because of two important reasons:
A. All of our users login with the 'same' exact login name/ password as it
is embedded in the desktop application exe file. So what we have here is
same login name/ password and different IP for each user
B. As mentioned before the server have 100 Mbps, my thoughts ('at first')
was that I wanted each user to get 'for example' maximum speed then 'all of
them' will have the same 10 MBps. But I never imagined that the connection
speed 100 or 10 will be (divided) on the number of users logged in, meaning
I couldn't imagine that when I drop the speed to 10 MBPs for user A then all
users will have this speed divided on the number of users logged into the
server (and this is of course due to my network basics ignorance :( )
So my question now is...is it possible in the first place that 'each user'
will get the same 10Mbps despite of the number of users connected to the
server (please excuse my network ignorance here if what I say seems
impossible)?
Now if this will not be possible, is it possible that I simply limit the
usage of the server for browsing html / html files only and exclude any
downloads exe, mp3, ...etc? without putting any limitation on speed? If I
can do this then there might be no need to do the delay pools limitation in
the first place!
Thanks in advance for your time and efforts
Amos Jeffries-2 wrote:
>
> mikewest09 wrote:
>> Hi All,
>>
>> Thanks Amos and Chris for your time and help, I have few more questions
>> so I
>> will appreciate it so much if you can please take a look at them
>>
>> A. As a start to avoid confusion here are some facts regarding our server
>> and users
>> - All of our users connect to our proxy (SQUID) through an application.
>> This
>> desktop application contains built-in (User name / password) for
>> authenticating each of our users to use our SQUID proxy (installed on our
>> server) AND also contains the IP of our server (which have SQUId
>> installed
>> on it)
>>
>> - Our users come from many different locations around the world and many
>> of
>> them (most of them actually) don't have 'static Ip's so I can't setup
>> specific range of IP's
>>
>> - We have more than 100,000 registered users (beside those whom will
>> register in the future) from different locations around the world. Of
>> course
>> not all of them will login together at a specific moment, but many of
>> them
>> can be on the server at the same moment and this was one of the major
>> reasons why we need to regulate SQUID traffic usage
>>
>> B. Even though I have read info about classes 1,2,3,4 I can't understand
>> up
>> to this moment which will be better for us? So based on the info I just
>> provided in A, do you have any recommendations especially when taking in
>> consideration the amount of users we have
>
> Depends on what you consider the most stable detail. username or IP.
>
> Given that you said most users are on dynamic IPs I'd go with username.
> class 4. With all fields except the user one set to -1/-1
>
>>
>> C. delay_parameters 1 -1/-1 1310720/15728640
>
>>>>> isn't -1/-1 is for unlimited access? i mean can you please explain
>>>>> this
>>>>> rule so I can understand its limitations?
>
> "unlimited" here means only "unlimited by Squid" the external network
> settings still apply.
>
>
> If you think of it like a geographic hierarchy it makes more sense.
>
>
> class 1: for limiting the entire network HTTP bandwidth.
>
> class 2: for limiting individual machines
> + maybe entire network.
>
> class 3: for limiting individual machines
> + subgroups of machines
> + maybe entire network.
>
> class 4: for limiting on login usernames
> + maybe individual machines
> + maybe subgroups of machines
> + maybe entire network.
>
> class 5: for advanced custom controls.
>
>
> Lets take a small office LAN as the theoretical situation:
>
>
> With class 1 the office can have a 10Mbps pipe and a policy of
> allowing only 2Mbps for web browsing traffic "aggregate" (8Mbps
> dedicated to the phones or external services etc).
>
> Effects: any single office PC can reach 2Mbps alone, or four can each
> simultaneously do 256Kbps. The 2Mbps is parceled equally between all.
>
>
> With class 2 they can take their 2Mbps traffic policy "aggregate" and
> add a condition that no one office machine can use more than 256Kbps at
> once "individual".
>
> Effects: running 3 or less PC at once will leave some hundreds of Kbpps
> "wasted" empty, even as each PC can runs at their peak 256Kbps. Five or
> more PCs will start to degrade each others experience as the 2Mbps is
> parceled equally between all at 200Kbps each.
>
>
> With class 3 they can have two offices next to each other who are
> allowed 1Mbps each. So one machine in office A can use 1Mbps by itself
> "individual" while two others in office B are at 500Kbps sharing the
> other 1Mbps "network". The total bandwidth stays within the overall pipe
> limit "aggregate".
>
>
> For you with class 2 I set non-individual caps as unlimited. Which means
> that Squid trusts the underlying network to get the pipe limits and
> sharing right. Squid will try to use as much bandwidth as needed to
> supply the current clients.
>
> Looks liek class 4 might be better for you, in which case the aggregate,
> network, and individual fields would be unlimited and only the username
> one set to the 10Mbps/15MB values.
>
> Amos
> --
> Please be using
> Current Stable Squid 2.7.STABLE7 or 3.0.STABLE20
> Current Beta Squid 3.1.0.15
>
>
--
View this message in context: http://old.nabble.com/Squid-delay-pool-question-tp26631264p26654741.html
Sent from the Squid - Users mailing list archive at Nabble.com.
