#2478 TimeInterval intepreted as 0.001 seconds instead of 0.01 seconds (for informs)

linux
closed
None
5
2013-08-30
2013-07-19
No

snmpd version: 5.4.1~dfsg-12 (from Debian 5.0.9)
Architecture: i386 (i686)
Kernel: Linux 2.6.26-2-686 (SMP w/3 CPU cores)

I have set up an inform-sink in the snmpd config file.
Querying SNMP-TARGET-MIB::snmpTargetAddrTimeout shows, that the timeout is
set to 1000 (which is the default value in net-SNMP).
SNMP-TARGET-MIB::snmpTargetAddrTimeout is of type TimeInterval, which is
defined as a period of time measured in 0.01 seconds (see
http://net-snmp.sourceforge.net/docs/mibs/snmpTargetMIB.html#TimeInterval
for details)

I am assuming, that informs will be re-sent (if not acknowledged) every
10 seconds (1000*0.01 seconds = 10 seconds)

But in fact, the retry interval, which I observed was 1 second (instead
of the expected 10 seconds)

This means, that in this case, TimeInterval in interpreted as 0.001
seconds instead of 0.01 seconds.

Either the documentation on TimeInterval is wrong, or
snmpTargetAddrTimeout has an incorrect Datatype assigned or
snmpTargetAddrTimeout is implemented incorrectly

Discussion

  • Wes Hardaker

    Wes Hardaker - 2013-07-22

    It was an issue with internal numbers clashing with mib representations. The MIB code has been fixed to do the appropriate multiple/divide of 10 to report it correctly in the mib tables. See commit 78dac6e.

     
  • Wes Hardaker

    Wes Hardaker - 2013-07-22
    • status: open --> closed
    • assigned_to: Wes Hardaker
     
  • Leonardo Chiquitto

    Looks like the follow up patch (commit 390303059) reintroduced the problem.

    $ snmpwalk -v2c -c public localhost snmpTargetAddr | grep Timeout
    SNMP-TARGET-MIB::snmpTargetAddrTimeout.'internal0' = INTEGER: 1000

     

Log in to post a comment.

Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:





No, thanks