Skip to:
SNMP is UDP based, and a number of factors can cause an SNMP request to get lost(network congestion, busy agent, etc.)
It appears that the SnmpCollector doesn't actually do retries:
synchronized (handler) {session.send(out, handler);
try {handler.wait((long) ((peer.getRetries() + 1) *peer.getTimeout()));} catch (InterruptedException e) {Thread.currentThread().interrupt();}}
Instead it just waits longer for a response. I could be wrong on this.
If I'm right, I think we should actually retry the request instead of justwaiting longer. I will test this in unstable as well.
This is still an issue. The code is now in SnmpSession.java:
Too old, reopen if this issue still exist.
moving to 1.6.1 target milestone; only things left pending for 1.6.0 are blocker-level bugs
SNMP is UDP based, and a number of factors can cause an SNMP request to get lost
(network congestion, busy agent, etc.)
It appears that the SnmpCollector doesn't actually do retries:
synchronized (handler) {
session.send(out, handler);
try {
handler.wait((long) ((peer.getRetries() + 1) *
peer.getTimeout()));
} catch (InterruptedException e) {
Thread.currentThread().interrupt();
}
}
Instead it just waits longer for a response. I could be wrong on this.
If I'm right, I think we should actually retry the request instead of just
waiting longer. I will test this in unstable as well.