This chapter lists an assortment of common problems and their solutions.
The problems include the following:
The default values for the controls are NTSC-specific. If you are using a 625/50 system, set the following system control in file /usr/vtr/config/system-defaults/main:
vtr.main.timing_standard 625 |
This control sets the timing of the system, including the multiformat codec (MFCODEC) cards, to 625/50.
In the file /usr/vtr/config/device-defaults/MFCODEC, make sure the control vtr.media.video.output.timing is set to its default value, system.
If vvtr crashes, you might have exceeded the number of configured pollable semaphores. The log would appear as follows:
C 06-14:29:23.348170 1386 Failed opening semaphore file descriptor (sems open=147): No space left on device A 06-14:29:23.348985 1386 FATAL SYSTEM ERROR: SYSTEM CONFIGURATION ERROR |
To fix the problem, follow these steps:
Edit the file /var/sysgen/master.d/usema and increase USMAXDEVS, for example, to 1050.
Rebuild the kernel (autoconfig) and reboot the system to activate the change.
The system normally boots from the primary root plex, that is, root.data.0.0. If the primary plex becomes unavailable, you can either label the disks and swap master and slave, or you can use the following procedure to set up the system so that it can boot from the secondary root plex, for example, system.data.1.0.
Ensure that sash (the standalone shell) is installed in the volume header of the secondary plex.
From the System Maintenance Menu, choose Enter Command Monitor (5).
Display the PROM environment variables:
>> printenv SystemPartition=dksc(0,1,8) OSLoadPartition=dksc(0,1,0) root=dks0d1s0 ... |
The swap PROM environment variable (which is set below) is not displayed because it is not saved in NVRAM.
Reset the SystemPartition, OSLoadPartition, and root environment variables to the values of the disk partition that contains the alternate plex and the swap environment variable to have the value of the alternate swap partition. For example:
>> setenv SystemPartition dksc(0,2,8) >> setenv OSLoadPartition dksc(0,2,0) >> setenv root dks0d2s0 >> setenv swap /dev/dsk/dks0d2s1 |
Exit the Command Monitor and restart the system:
>> exit ... Option? 1 Starting up the system... ... |
It is beyond the scope of this book to describe the intricacies of the IRIX filesystem features and tools that provide optimal performance for the SGI Media Server. For detailed information, see the latest version of IRIX Admin: Disks and Filesystems.
Inside the /usr/vtr/adm/crash directory is a subdirectory for each program that crashes. If a program crashes, a directory is dynamically created with that program name. Inside that directory a core file is created containing the crash information.
When the SGI Media Server detects a program crash, it renames the core file to core.number, where number is the next incremental number. This scheme prevents core files from overwriting each other.
SGI Media Server crash files are saved to aid SGI support personnel in diagnosing server software errors. Since core files may be periodically purged, generate a stack trace for the crash files to capture the pertinent information. Use the following procedure to generate a stack trace:
Log in to the SGI Media Server as root.
Change the directory to the appropriate core file location, as shown in the following:
server 1# cd /usr/vtr/adm/crash/vvtr/core.number |
The value number is the core file directory of interest.
Run the dbx command on the core file as shown in the following:
server 2# dbx ./vvtr ./core . . . (dbx) |
The dbx utility prints its header information and then prompts for more input.
Enter the following dbx command sequence:
(dbx)record output /usr/tmp/core.number.txt (dbx)where (Press the space bar or Enter key as needed.) (dbx)quit |
Send the /usr/tmp/core.number.txt file to SGI support personnel.