TMF Operations and Recovery Guide (G06.26+)

Recovery Methods
HP NonStop TMF Operations and Recovery Guide—522417-003
7-3
Volume Recovery
recovery operation with its current memory allocation. If this happens, use the ALTER
PROCESS command to increase the EXTENDEDSEGSIZE attribute of the backout
process and then stop the process. TMF automatically restarts the process with the
larger extended segment allocation. Refer to the TMF Reference Manual for
instructions on using the ALTER PROCESS command and to the TMF Planning and
Configuration Guide for specific recommendations.
When an error prevents the backout process from undoing all of the changes made by
a transaction that was aborted, the backout process generates an event message.
This message describes the error and identifies the transaction: the transaction
remains in the aborting state. The backout process tries several times to abort the
transaction; if it fails, the transaction state changes to hung, and the backout process
stops trying to abort it and sends a message to the EMS log.
Resolve the problem that caused the backout failure; if the failure has resulted in hung
transactions, then reissue the ABORT TRANSACTION command to cause the backout
process to retry backing out those transactions. See Controlling Individual
Transactions on page 3-21 for cautionary information and instructions.
When the backout process reports an error, you should correct it as soon as possible.
Audit-trail files that contain incomplete transactions remain pinned in the audit trail, so
unresolved backout problems can eventually lead to a suspension of transaction
processing if the begin-transaction-disable threshold is reached. Also, the records
modified by the transaction remain locked.
Volume Recovery
The volume recovery process is automatic to your TMF system; although you do not
have to invoke it, it is helpful to understand how the process works so you can interact
with it in special circumstances.
Volume recovery repairs audited files on a data volume after a system or volume
failure, and allows the volume to be started for TMF processing. Volume recovery is
initiated when a START TMF or ENABLE DATAVOLS command is issued, or when the
volume is brought up after a system or volume failure. A data volume requires volume
recovery processing if one of the following conditions occurs:
•
Both the primary and the backup processors (CPUs) for a disk process fail.
•
Both the primary and the backup disk controllers fail.
•
The volume is specified in an SCF STOP DISK command. This command is
described in the SCF Reference Manual for G-Series RVUs.
•
The disk process fails in an attempt to generate audit records.
•
A disk volume loses power or fails in some other capacity.
•
The entire system fails.
•
A TMF crash occurs (see Responding to a TMF Crash on page 7-29).