HP Insight Control for Linux 6.0 Release Notes

2.6.6 Nagios nrpe daemon might fail to start on the CMS
The nrpe daemon might fail to start on the CMS.
Every time you run Options→Configure Insight Control for Linux Management Services,
verify that the nrpe daemon is running on the CMS:
# ps auxww | grep nrpe
If nrpe is not running, enter the following command to start it:
# /etc/init.d/nagios start_nrpe
2.6.7 Nagios enclosure monitoring might fail if enclosure has multiple OAs
Nagios enclosure monitoring might fail if the enclosure has more than one OA. Discovery of
both the ACTIVE and STANDBY OAs might result in an incorrect bay_number for the ACTIVE
OA. You can confirm this with the shownode config enclosures command, as follows:
# shownode config enclosures | grep -B3 ACTIVE
192.0.2.221:
IPV6_address:
bay_number:
currentRole: ACTIVE
If the command returns an empty or an incorrect bay_number for the ACTIVE OA, follow the
corrective action.
Corrective action: Delete the STANDBY OA in SIM and rerun Options→IC-Linux→Configure
Management Services.
2.6.8 Absent /dev/mcelog device file causes messages on virtual hosts and guests
On virtual hosts and virtual guests, the /usr/sbin/mcelog executable file might exist but the
corresponding /dev/mcelog device might not exist if mcelog is disabled in the kernel.
If so, error messages like the following will continually show up in the /var/log/messages
and /hptc_cluster/adm/logs/consolidated.log files, and Nagios will generate Alerts
for these error messages:
date time callisto_guest_2/callisto_guest_2 mcelog: Cannot mmap 0xf0000: Bad address
date time callisto_guest_1/callisto_guest_1 mcelog: Cannot mmap 0xf0000: Bad address
date time callisto_guest_2/callisto_guest_2 mcelog: Cannot open /dev/mcelog
date time callisto_guest_1/callisto_guest_1 mcelog: Cannot open /dev/mcelog
date time callisto/callisto last message repeated 2 times
Corrective action: If the /dev/mcelog device does not exist on the managed system, rename
the /usr/sbin/mcelog executable file on the managed system to /usr/sbin/mcelog.orig
so that it does not run.
2.6.9 Nagios may generate WARNING "data is stale" messages if RRDs become
corrupted
On rare occasions, the RRDs, where Nagios stores metric and sensor data, might become corrupted.
If the RRD file sizes for a specific metric (for example, load average) are not identical for all the
managed systems then the RRDs are corrupted.
For example, in the output below, note that two of the three RRD files are of equal size (71104),
but the other file is not (35944). Note also that the shownode metrics load command reports
data for only one of the three managed systems.
# ls -lt /hptc_cluster/rrd/avenrun/
total 188
-rw-r----- 1 nagios apache 35944 date time 00000019bb28b16c0000000300000001.rrd
-rw-rw-r-- 1 nagios apache 71104 date time 00000019bb28b16c0000000300000007.rrd
-rw-rw-r-- 1 nagios apache 71104 date time 00000019bb28b16c0000000300000008.rrd
# shownode metrics load
Timestamp |Node |1 min |5 min |15 min |# Running |# threads
---------------------------------------------------------------------------------------
date time |icelx1 |0.60 |0.44 |0.38 |1 |573
18 Notes and corrective actions