WARNING..........make sure your system is backed up before doing anything or applying patches.. If you have problems, you will need to restore..........and you don't want your users screaming.................

Journaled File System Log Size Issues

Another size-related issue is the size of the JFS log. In most instances, multiple journaled file systems use a common log configured to be 4MB in size. For example, after initial installation, all file systems within the root volume group use logical volume hd8 as a common JFS log. The default logical volume partition size is 4MB, and the default log size is one partition, therefore, the root volume group normally contains a 4MB JFS log. When file systems exceed 2GB or when the total amount of file system space using a single log exceeds 2GB, the default log size may not be sufficient. In either case, the log sizes should be scaled upward as the file system size increases. The JFS log is limited to a maximum size of 256MB.  

You can theoretically improve your performance by having additional jfslogs
that are on different disks.  If a filesystem has to wait to write to a jfslog, performance suffers.  If you have additional jfslogs on different drives and point your filesystems to them, the wait should be decreased. 

Error Description: JFS_LOG_WRAP Error ID: 061675CF

This error is logged if transactions written to a JFS log device caused the beginning of the log to be overwritten. The system will have crashed as a result. The solution is either to move filesystems that use that log to other JFS log devices to attempt to distribute the amount of JFS log activity, or to increase the size of the affected JFS log. Use the major/minor numbers in the error report to identify the log device in the /dev directory.

NOTE: To increase the size of a JFS log device, all filesystems that use the log must be unmounted. The logical volume that the log device exists on can then be extended with extendlv to the appropriate size ( up to a maximum of 256meg). Afterwards, run logform on the log device to format it to use all of the space in the logical volume.

Error Description: Error ID: CF71B5B3

This error is logged if transactions written to a JFS log device are reaching a threshold that may result in a JFS_LOG_WRAP condition. This entry, which is intended as a warning, is logged if this condition occurs more than 10 times in a one-hour period. Use the solution given for JFS_LOG_WRAP to alleviate or eliminate the condition.

How do I set up a new jfslog?

Unmount the filesystem before you try to point the log to the new filesystem.
If a JFS log does not already exist on the new
volume group, then create one by using the 'mklv'
and 'logform' commands.

OR DO IN SMIT calling it jfslog for type............ but choosing
one of the disks that the filesystem is on......... smitty , then physical and logical storage, logical volume manager. Logical volumes, add a logical volume
Add a Logical Volume

Type or select values in entry fields.
Press Enter AFTER making all desired changes.
-----------------------------

 Logical volume NAME   []  (you can leave blank)
 * VOLUME GROUP name       [datavg] 
  (fill in your volume group with f4)
 * Number of LOGICAL PARTITIONS     [1]   # 
 ( one or two depending on pp size)

 PHYSICAL VOLUME names        [hdisk5] 
   (actually choose the volume name with the f4)

 Logical volume TYPE         [jfslog ] 
    (jfslog is the type)

 POSITION on physical volume middle   []+
 RANGE of physical volumes minimum  []+
 MAXIMUM NUMBER of PHYSICAL VOLUMES [] #
 Number of COPIES of each logical partition [1] +
 Mirror Write Consistency?    [yes] +
 Allocate each logical partition copy   [yes] +  
  -------------------------------------      

For example:

mklv -t jfslog -y datavg 1
(where "datavg" is the name of the users volume group.)
This will return a new JFS log lv such as "loglv01" then run logform on it.

logform loglv01

Change the file system to reference the new log device with the 'chfs' command.
For example:

chfs -a log=/dev/loglv01 /myfilesystem
(where "myfilesystem" is the name of the users file system.)
(you can then check in the /etc/filesystem file to be sure it has been changed) and also do a savebase from the command line.

You can fsck the unmounted filesystem to be sure all is correct...........

Mount the file system

For example:
mount /myfilesystem
(where "myfilesystem" is the name of the users file system.)

(slg03/2002)