HP NonStop RDF, Management Manual

Page: 249 / 478

Cpus
The eighth column specifies the CPUs in which each process pair is running.
Error
The final column is used for several purposes. For all RDF processes it is usually blank, which
indicates the process is running normally and without any error condition. The following displays
can also be reported.
****
The specific process has encountered a serious error. You should examine the event log to
see what happened. If the error condition is cleared, for example, an updater reporting a file
system error 122, then the asterisks are cleared. If the error condition is not corrected, the
asterisks continue to appear until the situation is corrected. If an updater encounters an
unexpected error 1, 10, 11, 71, or some other condition, this could indicate that your backup
database is no longer in synchronization with the primary system's database. If you stop
RDF and then restart it, any asterisks previously reported are cleared until a process encounters
a new unexpected error condition.
sync
During an online database synchronization operation, an updater reports "sync" in this
column until it has caught up and has gone past the point when the STOP SYNCH command
was issued.
undo
During a takeover or stop-update-to-time operation, after the update has finished its forward
moving redo pass, it reports "undo" when it performs its reverse moving undo pass in order
to back out any updates for transactions that need to be undone. In a takeover operation for
an RDF network, "undo" indicates the first undo pass when local transactions are undone
because their outcomes are unknown.
net
During an RDF network takeover operation, there are three possible undo phases: local undo
(marked "undo" as stated above), file-undo (not marked because this highly rare and typically
only lasts a few seconds), and network undo. The latter is reflected as "net", and this undo
pass backs out transactions for which RDF has a commit record in this system, but a commit
is missing on another backup system in the RDF takeover. See
“Network Transactions”
(page 295) for the details of an RDF network takeover operation.
For more information on critical errors, you can scan the EMS collectors on the primary and
backup systems:
• The EMS collector on the primary system contains log messages for the extractor and monitor
processes.
• The EMS collector on the backup system contains log messages for the receiver, purger, and
all updater processes.
Special Messages
If you issue the STATUS RDF command while an RDF TAKEOVER operation is in progress,
RDF displays the current state as “TAKEOVER IN PROGRESS”, as seen in the second example
above.
If you issue the STATUS RDF command after an RDF TAKEOVER operation has completed,
RDFCOM displays this message instead of the usual status display:
STATUS RDF (\RDF04 -> \RDF06) is NOT running
An RDF TAKEOVER has completed
Safe MAT position is SNO 1, RBA 87876660
MAT position for File Recovery: SNO 1, RBA 87876740
RDFCOM Commands 249