System Management Release Notes
4.30 Recovering From System Hangs or Crashes (Integrity servers Only)
If XDELTA is loaded, pressing Ctrl/P causes the system to enter XDELTA. The
system displays the instruction pointer and the current instruction. You can force
a crash from XDELTA by entering ;C, as shown in the following example:
$
Console Brk at 8068AD40
8068AD40! add r16 = r24, r16 ;; (New IPL = 3)
;C
If XDELTA is not loaded, pressing Ctrl/P a second time causes the system to
respond with the prompt ‘‘Crash? (Y/N)’’. Entering Y causes the system to crash.
Entering any other character has no effect on the system.
4.31 DECdtm/XA with Oracle 8i and 9i (Alpha Only)
V7.3-2
When you are using DECdtm/XA to coordinate transactions with the Oracle 8i/9i
XA Compliant Resource Manager (RM), do not use the dynamic registration XA
switch (xaoswd). Version 9.0.1.0.0 of the Oracle shareable library that supports
dynamic registration does not work. Always use the static registration XA switch
(xaosw) to bind the Oracle RM to the DECdtm/XA Veneer.
The DECdtm/XA V2.1 Gateway now has clusterwide transaction recovery support.
Transactions from applications that use a clusterwide DECdtm Gateway Domain
Log can now be recovered from any single-node failure. Gateway servers running
on the remaining cluster nodes can initiate the transaction recovery process on
behalf of the failed node.
4.32 Device Unit Number Increased
V8.2
In the past, OpenVMS would never create more than 10,000 cloned device units,
and unit numbers would wrap after 9999. This had become a limitation for some
devices, such as mailboxes or TCP/IP sockets.
Starting with OpenVMS Version 7.3-2, OpenVMS will create up to 32,767 devices
if the DEV$V_NNM bit is clear in UCB$L_DEVCHAR2 and if bit 2 is clear in the
DEVICE_NAMING system parameter. This does not require any device driver
change.
However, programs and command procedures that are coded to assume a
maximum device number of 9999 may need to be modified.
4.33 EDIT/FDL: Fixing Recommended Bucket Size
V7.3
Prior to OpenVMS Version 7.3, when running EDIT/FDL, the calculated bucket
sizes were always rounded up to the closest disk-cluster boundary, with a
maximum bucket size of 63. This could cause problems when the disk-cluster
size was large, but the ‘‘natural’’ bucket size for the file was small, because
the bucket size was rounded up to a much larger value than required. Larger
bucket sizes increase record and bucket lock contention, and can seriously impact
performance.
4–24 System Management Release Notes
