Downed Host Detection is broken in 0.8.6g?

[alexb]

Hello,

I have a problem monitoring an SNMMP device that does not return OID .1.3.6.1.2.1.1.3.0 (system uptime). Cacti thinks that the device is down and refuses to proceed. I have “Downed Host Detection” set to “Ping” and “Ping Type” to ICMP ping. Here is what these parameters are in database:
mysql> select * from settings where name in ('availability_method', 'ping_method');
+---------------------+-------+
| name | value |
+---------------------+-------+
| availability_method | 3 |
| ping_method | 1 |
+---------------------+-------+

I am using a custom PHP script to poll the device and I can verify that SNMP returns data using my script and snmpwalk. Oh, and pooler writes to log:
01/15/2006 10:17:06 PM - CMDPHP: Poller[0] Host[2] SNMP: Host did not respond to SNMP

tcpdump on target shows snmpget to get uptime no matter what I try:
22:19:31.147648 192.168.200.21.snmp > 192.168.201.22.46746: { SNMPv1 C=public@AdminServer { GetResponse(27) R=948005368 genErr@1 .1.3.6.1.4.1.140.625=[|snmp]} } (DF) (ttl 64, id 1438, len 88)



Any clues, please?

Thanks,
Alex