[rrd-users] Counter reset to zero initialy and over hours, no rrd updates

[Gerhard Ecaroh Froehlich]

Hello,

i am monitoring Oracle DB. Unfortunately counters will be initialy "zero" when
creating the database and gets reset every night to "zero" by restarting the
databases too. For hours there will be no user request (04:00 - 09:00), therefor
after reset to "zero" the snmp value will be "zero" till the first user request
comes in the morning.

In the below shown listing i have a snmp-ora-cmr-scms.rrd who was created, but
never got an update unlike snmp-ora-chr-scms.rrd which works fine. "snmpwalk"
for the updated rrd results in real value, for the not updated i get a "zero"
cause this *is* the real *good* value which may only will increase if theres a
bad configuration on oracle db.

So i have NaN#S and UNKN for a value, which normaly indicates a correct
configured system. Do i have the wrong datatype "COUNTER"? For Oracle OID in
this case it should be Counter32.

Anybody knows how to deal with long time "zero" and initialy "zero" counters?
How can i avoid having NaN caused by zeros. 

Another question: By the nature of the oracle answer some values can be "zero".
For instance, a Ratio could result in a timespan of values like "calls /
commits", here maybe "23456 / 0". But dividing through zero is no good idea.
What can i do in this case?

Any help is appreciated.


	bye, Ecaroh


--- Listings ---

# State of rrds
monhost:root,29% date
Fri Aug 23 12:10:51 MEST 2002
monhost:root,30% ls -ltr
total 926
-rw-r--r--   1 root bin      45800 Aug 23 12:08 snmp-ora-chr-scms.rrd
-rw-r--r--   1 root bin      30768 Aug 22 17:40 snmp-ora-cmr-scms.rrd

# Command for target snmp-ora-chr-scms.rrd
monhost:root,31% ./snmpwalk -p 161 oradb secret .1.3.6.1.4.1.111.4.1.1.1.2.2
enterprises.111.4.1.1.1.2.2 = Counter32: 43716


# Command for target snmp-ora-cmr-scms.rrd
monhost:root,32% ./snmpwalk -p 161 oradb secret .1.3.6.1.4.1.111.4.1.4.1.5.10.6
enterprises.111.4.1.4.1.5.10.6 = Counter32: 0
# The returnvalue "0" is the expected and wished *good* value!

# Dump for snmp-ora-cmr-scms.rrd
<!-- Round Robin Database Dump -->
<rrd>
        <version> 0001 </version>
        <step> 300 </step> <!-- Seconds -->
        <lastupdate> 1030030517 </lastupdate> <!-- 2002-08-22 17:35:17 MET DST -
->

        <ds>
                <name> lcr </name>
                <type> COUNTER </type>
                <minimal_heartbeat> 600 </minimal_heartbeat>
                <min> 0.0000000000e+00 </min>
                <max> NaN </max>

                <!-- PDP Status -->
                <last_ds> UNKN </last_ds>
                <value> 0.0000000000e+00 </value>
                <unknown_sec> 17 </unknown_sec>
        </ds>

        <ds>
                <name> lcp </name>
                <type> COUNTER </type>
                <minimal_heartbeat> 600 </minimal_heartbeat>
                <min> 0.0000000000e+00 </min>
                <max> NaN </max>

                <!-- PDP Status -->
                <last_ds> UNKN </last_ds>
                <value> 0.0000000000e+00 </value>
                <unknown_sec> 17 </unknown_sec>
        </ds>

<!-- Round Robin Archives -->
        <rra>
                <cf> AVERAGE </cf>
                <pdp_per_row> 1 </pdp_per_row> <!-- 300 seconds -->
                <xff> 1.0000000000e-01 </xff>

                <cdp_prep>
                        <ds><value> NaN </value>  <unknown_datapoints> 0 </unkno
wn_datapoints></ds>

--

--
Unsubscribe mailto:rrd-users-request at list.ee.ethz.ch?subject=unsubscribe
Help        mailto:rrd-users-request at list.ee.ethz.ch?subject=help
Archive     http://www.ee.ethz.ch/~slist/rrd-users
WebAdmin    http://www.ee.ethz.ch/~slist/lsg2.cgi