OK, my use-case is as follows.
I have a sensor that ingests large amounts of data (parsing network logs
/ netflow, passiveDNS, etc). This sensor stores the data in its own
format, with the timestamp being accurate to the second. If I were to
build an interface that responds to TAXII queries for the collector in
front of the sensor, and I were to query this data, I'd be dealing with
large data-sets and I'd want to paginate it.
Unless I set a limit that will probably blow through my memory
limitations, I have no other TAXII-compliant way to paginate the data in
sane chunks without either losing some data (anything beyond the memory
limit's envelope for a given second) or without blowing through my
memory limits.
Best regards,
Andras
On 09.09.19 16:00, Bret Jordan wrote:
> Andras,
>
> Thanks for the question. TAXII should work well for this use case. I
> do not see why it would not. Please keep in mind that the limits we
> were talking about are optional. So a server / sensor may have no limit
> which lets you pull all records at once.
>
> The sensor can dynamically add / figure out the date-added values how
> ever it needs to do so. So I am not sure why this would not work. Can
> you help me understand why you think it will not work? Or does this
> solve your concerns?
