5    Verifying the Installation with the clu_ivp Utility

After all systems are installed or upgraded, use the cluster installation verification program, clu_ivp, to detect configuration errors. The clu_ivp utility can test the installation of the Production Server, Available Server, or MEMORY CHANNEL products; the type of tests performed depend on the product.

Note

Because the utility tests only the system on which it is run, you must run clu_ivp on each system to verify the basic configuration of a Production Server cluster or an Available Server configuration.

As part of checking the installation configuration, clu_ivp calls two other utilities: cnxshow and drd_ivp. See clu_ivp(8), cnxshow(8), and drd_ivp(8) for more detailed information on each utility. Use these utilities and the information in this chapter when verifying an installation. These utilities do not modify the system in any way. You can run them at any time to diagnose problems.

By default, the clu_ivp utility displays error conditions only. For example:

# /usr/sbin/clu_ivp
Cluster Installation Verification Procedure (IVP)
ERROR: asemgr has not been run to specify the member names.
Resolution: run 'asemgr' to specify the member names.
ERROR: drd_ivp validation tests failed.
Resolution: run 'drd_ivp -p -v -c' for more details.
 
Installation verification failed, 2 errors detected.

When an error is detected, the clu_ivp utility suggests corrective action. In some cases, the error reported by the clu_ivp utility is the symptom of another problem. Read all the error messages generated by the clu_ivp utility before attempting to correct problems. When the corrective action suggested by the clu_ivp utility does not solve the problem, examine the system's error log files and console output for additional clues.

Note

Following a first-time Production Server installation before any distributed raw disk (DRD) services are configured, you might see drd_ivp errors. With the exception of drd_ivp validation tests failed type messages, these errors are generally benign.

For more informative output, use the clu_ivp -v (verbose) option. In addition to reporting error conditions, the utility displays confirmation of each verification check as it is performed.

When run in verbose mode, the clu_ivp utility looks for configured controllers and SCSI buses that are eligible for sharing. It finds controllers that have records in the /sys/conf/HOSTNAME file; these records are created by booting the /genvmunix kernel and running the doconfig program. The following sample output from clu_ivp shows information on SCSI controllers and devices:

Found these supported SCSI controllers:
------------------------------
  pza0 is scsi2
  pza1 is scsi3
  pza2 is scsi4
Scanning EDT.........
Found these devices on bus scsi2:
  device              vendor   drive            firmware
  ------              ------   -----            --------
  rrz18c              DEC      RZ26N            0616
  rrz20c              DEC      RZ28             442D
Found these devices on bus scsi3:
  device              vendor   drive            firmware
  ------              ------   -----            --------
  rrz27c              DEC      RZ28             D41C
  rrz29c              DEC      RZ26             392A
Found these devices on bus scsi4:
  device              vendor   drive            firmware
  ------              ------   -----            --------
  rrz34c              DEC      RZ28B            0006
  rrz36c              DEC      RZ26N            0616
 
 

If the clu_ivp utility cannot find any controllers, one possibility is that the doconfig program was not run under /genvmunix since those controllers were installed. Because the clu_ivp utility scans the shared SCSI buses directly to detect disk devices connected to these buses, there is no need to run the doconfig program after adding a disk to a shared SCSI bus to make the disk visible to the clu_ivp utility.

To check that all members in an ASE have the same view of shared devices (same bus numbers, same device numbers), run clu_ivp -v on each system and compare the results.

You can also use the file command to verify that other systems connected to the shared bus see the same devices on the shared buses. In the following example, the output of the file command on system clu14 indicates that device rz18 is on the same bus (number 2) as on system clu13:

# file /dev/rrz18c
/dev/rrz18c: character special (8/34818) SCSI #2 RZ26N disk \
             #144 (SCSI ID #2) (SCSI LUN #0)
 

Note that because of device reservations, the file command reports errors after an ASE service is started on a disk if the command is invoked on a system that is not running the service.

The clu_ivp utility cannot tell whether your SCSI cabling is correct. If you have any questions about cabling requirements, see the TruCluster Software Products Hardware Configuration manual.