[erlang-questions] os_mon_mib error

[Edwin Fine]

Thanks Martin!

It makes sense - the system has 8 GB of RAM (8386678784). Maybe what I will
do in the meantime is extend the MIB as you suggest just so I can move
forward. I'd rather not truncate the value since it is accurate.

Also, there are two values in the os_mon MIB that are incompletely defined:
loadAlarmsGroup and diskAlarmsGroup. I commented them out to avoid the error
messages.

Do I need to put in a bug report somewhere?

Regards,
Edwin


On Wed, May 28, 2008 at 2:19 AM, Martin Bjorklund <mbj@REDACTED> wrote:

> Hi,
>
> "Edwin Fine" <erlang-questions_efine@REDACTED> wrote:
> > Hi all,
> >
> > I am learning to use OTP snmp. I've done an snmp:config() for a v3 agent,
> > which is running ok, it seems. However, when I do an snmpwalk, I get a
> > wrongValue error. I'm running Ubuntu Feisty x86_64 using Erlang/OTP
> R12B-2.
> > This only happens if I do an os_mon_mib:load(snmp_master_agent).
> >
> > What am I doing wrong?
>
> This is a bug in os_mon.
>
> > $ snmpwalk -v 3 -l authPriv -m ALL -u initial -a MD5 -A **** -x DES -X
> ****
> > 127.0.0.1:4000 otp
> > ... snip ...
> > OTP-MIB::applVsn.1.7 = STRING: 2.12.2
> > OTP-OS-MON-MIB::loadMemorySystemWatermark.0 = INTEGER: 80
> > OTP-OS-MON-MIB::loadMemoryErlProcWatermark.0 = INTEGER: 5
> > Error in packet.
> > Reason: (genError) A general failure occured
> > Failed object: OTP-OS-MON-MIB::loadMemoryErlProcWatermark.0
>
> This tells us that doing a GET-NEXT on loadMemoryErlProcWatermark.0
> fails.  The object after this one in the MIB is
> loadSystemTotalMemory.
>
> > The agent error is:
> >
> > User error: Got 8386678784 from {os_mon_mib,load_table,[]}. Using
> wrongValue
>
> This tells us that when the agent called the instrumentation function
> for loadSystemTotalMemory (which is the function load_table in module
> os_mon_mib), it returned 8386678784, which the agent says is a bad
> value.
>
> The reason for the wrongValue is that loadSystemTotalMemory is a
> Gauge32, and 8386678784 is not a 32-bit integer.  And this is since
> you run on a 64-bit platform.
>
> The MIB could be extended with an object of type CounterBasedGauge64
> from the HCNUM-TC MIB (RFC2856), or the implementation would have to
> truncate this value.  In either case - bug in os_mon.
>
>
> /martin
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://erlang.org/pipermail/erlang-questions/attachments/20080528/935b05e1/attachment.htm>