Using SNMP traps for the TS7650G (Gateway), ProtecTIER V3.4.3

About this task

In the event of hardware or software degradation or failure, ProtecTIER® systems which are configured to use Simple Network Management Protocol (SNMP) can send a problem notification to designated recipients. SNMP notifications, or traps, can be sent even if the ProtecTIER Manager interface is unavailable.

To use SNMP traps you need the following items:

SNMP trap receiver software installed on an SNMP trap server. Follow the instructions from the manufacturer to install and configure the SNMP trap receiver software.
The file name and location of the management information base (MIB) file for the SNMP trap receiver. On the ProtecTIER server, the file name is: IBM-TS7600-SNMP-MIBV2.mib located in: /usr/share/snmp/mibs. The full path is: /usr/share/snmp/mibs/IBM-TS7600-SNMP-MIBV2.mib.
The IBM-TS7600-SNMP-MIBV2.mib file needs to be copied onto the SNMP trap receiver and the trap receiver software must point to the directory location of the MIB file for translation of the trap messaging.
SNMP trapping enabled on one or more of your ProtecTIER servers. For 3958 DD4 servers, see the IBM® ProtecTIER User's Guide for VTL Systems, GA32-0922.

The ProtecTIER servers have the following improvements in SNMP support.

ProtecTIER software events that send specific notifications based on the error that occurred.
ProtecTIER hardware events that trigger specific notifications are based on the error that occurred, such as a CPU event or power event.
Send enough detailed information with the SNMP notification so that you can understand the problem. The ProtecTIER Manager Configuration menu gives you the option to filter SNMP traps based on severity.
- Error-level severities can be filtered by:
  - Error
  - Warning
  - Information
  - Software error categories include:
    - VTL
    - Replication
    - Repository storage
    - Cluster
    - System
  - Hardware error categories include:
    - CPU memory module
    - Cooling module (fan)
    - Internal boot drives
    - Ethernet cards
    - Power supplies
    - RAID card
    - RAID battery
    - Front end adapter, if VTL enabled
    - General server errors
    - General network errors
    - Ethernet switch, if cluster enabled with SMC switch (TS7650 or TS7650G only)
    - Network power switch, if cluster enabled with new network power switch (TS7650 or TS7650G only)
    - Back end adapter (TS7650 or TS7650G only)
    - Disk controller (TS7650 or TS7650G only)
    - Disk expansion (TS7650 or TS7650G only)
    - SAS expander
    - SATA hard disk drives
- Warning-level severity includes:
  - Replication warnings
  - VTL warnings
  - Capacity warnings
  - RAS warnings
- Information-level severity includes:
  - VTL configuration change events
  - Replication events
SNMP in ProtecTIER version 3.1 or later supports threshold monitoring and allows the user to specify thresholds for the following system runtime behavior:
- Repository space issues
  - Nominal capacity
  - Physical capacity
- There are two threshold levels a user can set:
  - Information level: a trap is sent when the repository regains free space and rises about the information level.
  - Warning level: a trap is sent when the free space in the repository falls below the warning level
- Going below the informational threshold issues an SNMP trap only if the warning threshold has been crossed. This method is to ensure that the user is not flooded with alerts when the normal operation crosses the low water mark threshold frequently.
- Capacity thresholds can be set specifying % from the repository or specifying space (GBs).
Using an IBM-registered management information base (MIB) file.
- The MIB file is implemented in a tree structure and has a unique OID for each message supported.
- The MIB file ships on the ProtecTIER server.
Provide reporting to the network management application software.
Improved communication options:
- SNMP traps are sent through the customer network (eth0) by using the UDP protocol.
- By default, port 162 is used and up to five destinations are supported.
- Customers can optionally select a different port for SNMP traffic by using the ProtecTIER Manager Configuration menu.

On systems configured to use SNMP traps, an agent monitors the ProtecTIER server and reports fault information to a network management application. Periodically the data is sent to the designated SNMP server in the form of an SNMP trap report, a portion of which is shown in Figure 1. SNMP trap reports allow you to receive hardware or software fault notifications whether or not you have access to the ProtecTIER Manager interface. The display format of the trap report varies between different trap receiver software applications. Your trap report might not look exactly like the following example.

Figure 1. SNMP trap report