fujiwara(_at_)jprs(_dot_)co(_dot_)jp wrote:
Hello, I'm Kazunori Fujiwara, JPRS.
I was in MARID bof in Seoul, IETF59.
I wondered why only smtp was considered in this BOF.
(But I could not comment because my English problem.)
Because I got many SPAM SIP calls.
I set SIP resource record in DNS.
and I made a presentation
with this SIP URI as a example for ENUM in JAPAN.
then, some person's mistake or DoS caused many SPAM calls.
Can you provide some more information on this? What kind of call were
they? Did they have an origin number? Were they recorded or computer
generated?
/*
sorry, I'm SIP newbie.
But current SIP implementations (or SIP protocol) have no protection
mechanism without SIPS or IP address restriction, I think.
*/
So, I think we need rough protection mechanism for any protocols.
My idea is
- Service origin addresses is small number and finite.
- SRV resource record is a good idea.
/* MX RR may equal to _smtp._tcp.DOMAIN SRV 0 0 25 MXhost. */
- I propose SRVORIGIN resource record for any protocols.
_smtp._tcp.DOMAIN IN SRVORIGIN hostname.
hostname. IN A ....
hostname. IN AAAA ....
If we have 130 service origin IPv4 addresses,
we write 10 SRVORIGIN RRs and each hostname has 13 A RRs.
Is such idea OK?
I am not a SIP expert, but you might want to run this by the SIP groups
also.
Yakov