| Age | Commit message (Collapse) | Author |
|---|
|
snmp requests are now packaged into pseudo-continuations
to allow for being dispatched to seperate processes;
lightly tested for interoperability with NetSNMP, but
doesn't implement the complete set of AgentX messages
while here, clean up return types of mps_get* functions,
and make smi_insert refuse to insert duplicate OIDs
okay benno@ reyk@
|
|
being-standardized <endian.h>
ok deraadt@ millert@ beck@
|
|
AgentX notifications are the only portion implemented right now;
get in the tree to flesh out the remainder.
ok reyk@ benno@
|
|
|
|
Traps are still sent via SNMPv2 protocol. They can neither be
authenticated nor encrypted. - Transport mode is still UDP. Not
additional transport subsystems were added. - Only the User-based
Security Model (USM, RFC3414) is supported. View-Based Access Control
(VACM, RFC3415) is not included. - Configuration is described in
snmpd.conf(5).
This diff includes a few minor changes to ber.c/h adding a necessary
callback for the USM HMAC calculation and merging it with other
minor changes from ldapd's ber code.
From Gerhard Roth at genua
ok claudio@ reyk@
|
|
|
|
restricted connection will reject any commands except snmp traps.
|
|
|
|
|
|
|
|
applications will be able to send imsgs defining the trap to snmpd and
the daemon will do the ASN.1/BER encoding before sending traps to the
registered receivers. there are two advantages of this approach that
a) the applications do not need to handle any ASN.1/BER encoding and
b) snmpd will provide a central interface to define trap receivers.
discussed with thib and others
|
|
|
|
for OpenBSD. SNMP is a necessary evil. This is work in progress, don't
expect too much from it yet.
ok deraadt@
|
