[ndnSIM] [EXT]Re: customise PIT size

[Junxiao Shi]

Hi Sai

The forwarding plane in ndnSIM does not enforce PIT capacity limit.
Every unsatisfied Interest will stay in the PIT until InterestLifetime has
elapsed.
If the host machine has insufficient RAM to accommodate these PIT entries,
the simulation crashes.

If a forwarder is enforcing PIT capacity limit, having too many unsatisfied
Interests would either cause new Interests to be rejected or cause old PIT
entries to be evicted, depending on the design of the forwarding plane.

Yours, Junxiao

On Fri, Jun 24, 2022, 18:45 Sai Gautam Mandapati <saigautam96 at gmail.com>
wrote:

> *External Email*
>
> So, suppose my network and producer are not able to statisfy the consumer
> interest. Unsatisfied interests increase the PIT size. In that case would
> flooding be even possible?
>
>
> On Sat, 25 Jun 2022 at 8:29 am, Junxiao Shi <shijunxiao at email.arizona.edu>
> wrote:
>
>> Hi Sai
>>
>> If InterestLifetime is absent in the packet, it's equivalent to setting
>> it to 4000ms.
>>
>> If the consumer transmits Interests more frequently,
>> * If the network & producer can satisfy these Interests quickly, PIT size
>> remains small.
>> * If some Interests are unsatisfied, PIT size increases.
>>
>> Yours, Junxiao
>>
>> On Fri, Jun 24, 2022, 18:23 Sai Gautam Mandapati <saigautam96 at gmail.com>
>> wrote:
>>
>>> *External Email*
>>> Hi Junxiao,
>>>
>>> Thanks for your reply. If that is the case if I remove the lifetime of
>>> interest attribute and increase the frequency of packets with random
>>> content would I be able to achieve PIT flooding?
>>>
>>> Thanks,
>>> Sai.
>>>
>>> On Sat, 25 Jun 2022 at 7:33 am, Junxiao Shi <
>>> shijunxiao at email.arizona.edu> wrote:
>>>
>>>> Hi Sai
>>>>
>>>> As defined in NDN protocol, the PIT entry is deleted under these two
>>>> conditions, whichever occurs earlier:
>>>>
>>>>    - Incoming Data packet satisfies the PIT entry.
>>>>    - InterestLifetime has elapsed since the last time an Interest
>>>>    matching the PIT entry is received.
>>>>
>>>> Hence,
>>>>
>>>>    - If the Interests are being satisfied by Data, InterestLifetime
>>>>    does not affect PIT size much.
>>>>    - If the Interests are not being satisfied, higher InterestLifetime
>>>>    results in larger PIT size.
>>>>
>>>>
>>>> Yours, Junxiao
>>>>
>>>> On Sun, Jun 19, 2022 at 9:12 AM Sai Gautam Mandapati <
>>>> saigautam96 at gmail.com> wrote:
>>>>
>>>>> *External Email*
>>>>>
>>>> Hi everyone,
>>>>>
>>>>> From the previous PIT size-related email lists. I found that from
>>>>> ndnsim 2.0, nfd is used for PIT and custom fields for PIT have been removed
>>>>> further. I see that interest lifetime is the same as PIT entry time. In my
>>>>> simulation, I set the Lifetime, but I don't know how exactly this is
>>>>> influencing the PIT. Is there any way I can see what's happening after the
>>>>> interest LIfeTime for PIT? Please help me out. Thank you.
>>>>>
>>>>> --
>>>>> Sai Gautam
>>>>>
>>>> --
>>> Sai Gautam
>>>
>> --
> Sai Gautam
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.lists.cs.ucla.edu/pipermail/ndnsim/attachments/20220625/3db52cff/attachment.html>