Hardware Sentry KM for PATROL

Release Notes for v1.8.02

Home  Previous  Next

What's New

New Supported Platforms

Dell PowerVault TL2000 and TL4000 tape libraries
IBM TS3100 tape libraries
IBM TS3200 tape libraries
IBM v7000 Disk Arrays
Fujitsu-Siemens PRIMERGY servers running VMware

New Macros for Alert Actions

The following new macros can now be used in Alert Actions:
%{SYSTEM_NAME}: Monitored system identifier (i.e. object ID of the MS_HW_MAIN instance, same as the current %{HOSTNAME} which will remain for compatibility reasons)
%{SYSTEM_HOSTNAME}: IP address or full name as specified by the user when adding the new monitored system (i.e. MS_HW_MAIN/<id>/hostname)
%{SYSTEM_FQDN}: FQDN (fully qualified domain name) of the system, as exposed to BPPM (i.e. MS_HW_MAIN/<id>/MetaFQDN)
%{SYSTEM_TOKENID}: BPPM device Token-ID (i.e. MS_HW_MAIN/<id>/MetaTokenID)

New Configuration Variables

Configuration Variable



Alert after N Times value for connector status parameters.


Alert after N Times value for discrete parameters.


If set to 1, no MetaTokenID and no MetaFQDN information will be set in the PATROL namespace for any host.


Alert after N Times value for numeric parameters.


Alert after N Times value for Present parameters.


Path to the OpenSSH PrivateKeyFile.


A backup of the PATROL Agent configuration is now performed and saved in the $PATROL_HOME/config directory before a Reinitialization of the KM.
It is now possible to provide a private key and a passphrase when monitoring a remote host through SSH.
The Missing parameter has been replaced by a Present parameter to detect missing devices. Contrary to the Missing parameter, the Present parameter is always activated.

Changes and Enhancements

Improved Platforms Support

Brocade Network Advisor (SMI Agent): the KM now monitors Brocade SAN switches via Brocade's SMI Agent which provides information not available using SNMP, notably:
Ports in Virtual Fabrics
FCoE Ethernet Ports
Dell servers with iDRAC7 or later: The KM now supports Dell's new iDRAC7 management card, which provides full hardware information about Dell server.
Dell PowerEdge M1000e Chassis: The MS_HW_DellCMC.hdf connector now monitors the overall status of each blade installed in the chassis and makes possible the collect of the ambient temperature sensor in most recent versions of the Dell CMC firmware.
EMC Storage Systems: The KM has been modified to improved integration and monitoring of EMC Storage systems within the BMC ProactiveNet Performance Management environment.
Fujitsu PRIMEPOWER: The product provides enhanced support for Fujitsu PRIMEPOWER servers running Solaris. The “sun4us” hardware class is now supported as the “sun4u” class.
Hitachi HDS Disk Arrays: This version of the KM provides a new connector specifically designed and optimized for Hitachi HDS Disk Arrays. It adds new parameters and significantly reduces resources needed to monitor a Hitachi Disk Array. This connector replaces the SMI-S Compliant Disk Arrays connector which no longer needs to be used for Hitachi storage systems.
HP ProLiant servers: Invalid (usually zero) thresholds for temperature sensors are no longer reported. The KM now reports status parameter for temperature sensors to ensure alerts are still generated.
HP-UX: The KM now monitors HBAs in HP 9000 (PA-RISC) servers running HP-UX (link speed, link status and port status). Using the MS_HW_HPUXHBA.hdf connector requires root privileges.
IBM Blade Chassis - The connector for IBM Blade Chassis has been improved to:
add a status parameter for Fan Modules
collect an accurate Power Consumption
add monitoring of 3rd and 4th Fan Modules in chassis with 4 Fan modules
monitor second temperature sensor for BladeCenter HT chassis
IBM VIO Servers - The product:
had its MS_HW_IBMVIOS.hdf connector modified to no longer support unrestricted shell. To monitor an IBM VIO Server in an unrestricted shell, use the standard IBM AIX connectors
monitors Disk Controllers and  LUNs (AvailablePathcount)
provides an ErrorCount parameter for Physical Disks. Also, the Status parameter has been removed (No reliable physical disk status is available)
IBM DS6000/8000 Disk Arrays: The product provides improved monitoring for IBM DS6000/DS8000 Disk Arrays. The MS_HW_IBMDS68DiskArray.hdf connector adds new parameters, provides more details, better identifies the different components in the system and significantly reduces resources needed to monitor an IBM DS6000/DS8000.
SUN servers: The KM now monitors network ports on servers running Solaris 11 and network port's Device Instance ID when it is not the same as its Link Assigned Name. (e.g. the physical port device ID is e1000g0 and the ifconfig name is net0).
Oracle servers: The KM now provides support for Oracle servers (formerly known as Sun servers) through their ILOM card.
DataDirect Networks: The KM now monitors DataDirect Networks Storage Appliances.

Other Improvements

Network cards: the physical address of the device the network interface is connected to will now be displayed in the Inventory Report when available.
The content of the PATROL event triggered by default upon hardware failure has been modified to expose the system name and hostname of the monitored system. Warning: If another software leverages this PATROL event (BEM, BPPM, etc.), make sure the added information does not break the integration of the event (parsing issues notably).
The KM now supports unlimited debug file size. In previous versions, the debug file size was limited to 500 MB.
The ErrorPercent parameter of the MS_HW_NETWORK class now uses the ANYINST/ALL_INSTANCES mechanism. Thresholds are not instance specific anymore.
To improve performance, the MS_HW_BrocadeSwitchWBEM.hdf connector has been modified to activate unconditionally the following parameters in order to reduce the thresholds configuration and improve the customization of thresholds:

 In some cases, these parameters will therefore be activated even though they are not collected. They will stay OFFLINE.

To improve performance, the BandwidthUtilization parameter has been modified to rely on the ANYINST/ALL_INSTANCES mechanism unless necessary to do otherwise.
The overall performance of the monitoring of large Brocade environments from one agent has been greatly improved.
On HP ProLiant servers running VMware or Linux, each card is better identified with its interface name (eth0, vmnic0, etc.) and model.
SMI-S Compliant RAID Controllers:
the label of the disks was modified so that they are better identified.
the monitoring of battery backup units on RAID Controllers is now fully supported.
The product now supports the WMI version of the Fujitsu-Siemens ServerView agent. The SNMP version of the connectors has been updated to ensure they don’t activate when the WMI version is enabled.
StorageTek StreamLine Tape Library: the product now monitors the library's robotics system. Also, on tape drives, status values are now properly interpreted.
Dell Storage Manager Identifiers: The KM will now use Dell's Location (Nexus) codes to identify disks. The previously used SNMP identifiers are no longer reliable indicators of the true location of the physical disk and can change.
The possibility to set Alert After N Times values has been re-added for discrete, numeric and connector parameters. However, these settings apply to the entire KM and cannot be customized for each host anymore.
The monitoring of recent versions of Linux (RedHat Enterprise Linux 6 a.k.a. RHEL6, SuSe Linux Enterprise Server 11 SP2 a.k.a. SLES 11 SP2, etc.) has been improved to allow support of the new output format used by the latest version of the multipath utility.
The debug log has been improved to provide additional useful information. Also, when activated the debug mode now creates a debug file per host.

Fixed Issues

On IBM DS3000, DS4000 and DS5000 Series storage systems, physical disks with the status of "By-passed" are now discovered and an appropriate alarm is generated.
A problem related to the Java path prevented the version 1.8.01 of the KM from collecting data when Java is required on Linux/UNIX systems. This notably affected Telnet, SSH and WBEM-based collection.
The HP Insight Management Agent–Server connector (MS_HW_CpMgServNT.hdf) triggered temperature alerts due to invalid thresholds.
Physical disks were not discovered in VIO Server when monitored from a Linux system.
Modifications brought to Network Link Settings were not taken into account in version 1.8.01 of the KM.
Users could not modify the debug end time once the debug mode was turned on.
A problem related to the AWK script using octal notation for some special characters prevented the following connectors to operate properly on Linux and some Solaris systems:        

















Blade instances (for the MS_HW_BLADE class) were incorrectly attached to the agent device in BPPM instead of the appropriate MS_HW_MAIN device. As a result, the monitoring of Cisco SAN switches had objects spread between the device representing the Cisco SAN switch itself and the device representing the PATROL Agent in the BPPM Console.
Logical volumes were not properly discovered and monitored on Windows servers with a non-English character set (Chinese, Korean, Japanese, etc.).
Modifications of Intrusion Status settings were not taken into account in version 1.8.01. This issue is fixed in the 1.8.02 version.
In Cisco UCS blade chassis, the status of the blade servers was not properly reported when the blade had a specific user label configured.
The following issues on HP ProLiant servers running VMware or Linux have been fixed:
parameters for the network interfaces are now collected
the LinkStatus of the network cards is now properly collected
all network cards are now properly discovered
On IBM xSeries, the 6.3.2 version of the IBM Director agent reports voltage and temperature alert thresholds in a way that is inconsistent with the previous versions. The new connector takes into account this variation and support older and newer versions of the IBM Director Agent.
Cisco MD Sand Nexus Switches: Ports that were "Administratively Down" or "Out of Service" were previously reported as having a status of "Unknown". These ports will now report a status of "OK".
The VMWare ESX Disks (Storage) and VMware ESX Disks (IPMI) connectors are now automatically deactivated if the SMI-S Compliant RAID Controller connector applies to the monitored system.
Password used to connect to a WBEM server (SMI-S, cimserver, VMware) containing a special character, like ‘\’ (backslash), ‘<’ (lower than) or ‘>’ (greater than) were misinterpreted by the WBEM client.
Some Sun SPARC systems do not report the color of the LED. Without the LED color, the product was unable to set the appropriate alert thresholds. The KM now looks at the description of the LED to determine proper thresholds if the color is not available.
On EMC Clariion and VNX storage systems, failed (or disabled, or removed) storage processors and controllers are now properly reported as such.
HBAs are now discovered on VIO Server when monitored from a Linux or Solaris system.
Monitoring Sun/Oracle servers: when monitored through their ILOM card, voltage sensors no longer trigger false alarms. Also, when monitored through their ILOM card, the PredictedFailure parameter is now properly collected.
On HP BladeSystem c7000 chassis, some temperature sensors were not properly discovered.
The discovery would timeout when gathering information about the unconfigured HBAs in IBM AIX systems. As a result, no HBA were discovered at all.