HP OpenVMS Systems Documentation

This section describes how to manage Fast Path.

10.2.2.1 Fast Path System Parameters

There are three FAST_PATH system parameters:

• FAST_PATH
• FAST_PATH_PORTS
• IO_PREFER_CPUS

These parameters can be used to control Fast Path as follows:

FAST_PATH

FAST_PATH is a static system parameter that enables (1) or disables (0)the Fast Path performance features for all Fast Path-capable ports.

Starting in OpenVMS Version 7.2, Fast Path is enabled by default. InVersions 7.0 and 7.1, Fast Path was disabled by default.

FAST_PATH_PORTS

FAST_PATH_PORTS is a 32-bit mask. Once Fast Path has been enabled bysetting FAST_PATH to 1, FAST_PATH_PORTS can be used to selectivelydisable Fast Path for some specific adapter types.

The value of the FAST_PATH_PORTS system parameter is the sum of thevalues of the bits that have been set. Following is a description ofthe bit mask:

Bit	Mask	Description
0	00000001	0 = Fast Path is ENABLED for KZPBA ports when FAST_PATH is set to 1.
		1 = Fast Path is DISABLED for KZPBA ports.
1	00000002	0 = Fast Path is ENABLED for KGPSA ports when FAST_PATH is set to 1.
		1 = Fast Path is DISABLED for KGPSA ports.

The remaining bits are reserved for possible future adapter types.

The default setting for FAST_PATH_PORTS is 0, so neither the KZPBA northe KGPSA ports are disabled.

Note that CI drivers are not controlled by FAST_PATH_PORTS. Fast Pathfor CI is enabled and disabled exclusively by the FAST_PATH systemparameter.

IO_PREFER_CPUS

IO_PREFER_CPUS is a dynamic system parameter that controls the set ofCPUs available for use as Fast Path preferred CPUs.

IO_PREFER_CPUS is a CPU bit mask specifying the CPUs that are allowedto serve as preferred CPUs and thus can be assigned a Fast Path port.CPUs whose bit is set in the IO_PREFER_CPUS bit mask are enabled forFast Path port assignment. IO_PREFER_CPUS defaults to -1, whichspecifies that all CPUs are allowed to be assigned Fast Path ports.

You may want to disable the primary CPU from serving as a preferred CPUby clearing it's bit in IO_PREFER_CPUS. This will reserve the primaryfor use by non-Fast Path IO operations.

Changing the value of IO_PREFER_CPUS causes the FASTPATH_SERVER processto execute the automatic assignment algorithm that spreads Fast Pathports evenly among the new set of usable CPUs.

Following are the full set of commands used to identify and set apreferred CPU for a port.

DCL SHOW DEVICE/FULL or $GETDVI DVI$_PREFERRED_CPU

To identify the preferred CPU for any Fast Path-capable device whenFast Path is enabled, use the DCL command SHOW DEVICE/FULL to display -whether or not the device supports Fast Path - the current preferredCPU ID and, if set, the User Preferred CPU ID for a port or disk device.

Alternatively, the $GETDVI system service or the DCL F$GETDVI lexicalfunction will return the preferred CPU for a given device or file. The$GETDVI system service item code is DVI$_PREFERRED_CPU, and theF$GETDVI item code string argument is PREFERRED_CPU. The returnargument is a 32-bit CPU bit mask with a bit set indicating thepreferred CPU. A return argument containing a bit mask of zeroindicates that no preferred CPU exists, either because Fast Path isdisabled or the device is not a Fast Path-capable device. The returnargument serves as a CPU bit mask input argument to the$PROCESS_AFFINITY system service. The argument can be used to assign anapplication process to the optimal preferred CPU.

For an application seeking optimal Fast Path benefits, you can codeeach application process to identify and run on the preferred CPU wherethe majority of the process's I/O activity occurs.

A high-availability feature of OpenVMS Cluster Systems is thatdual-pathed devices automatically fail over to a secondary path, if theprimary path becomes inoperable. Because a Fast Path device could failover to another path or port, and thereby, to another preferred CPU, anapplication should occasionally reissue the $GETDVI in a timer threadto check that process assignment is optimal.

DCL SHOW CPU /FULL

You can use this DCL command to identify whether a CPU is enabled foruse as a preferred CPU, and the current set of ports assigned to thatCPU.

DCL SET /PREFERRED_CPU and /NOPREFERRED_CPU

These commands allow you to specify a CPU or a set of candidate CPUsfrom which the operating system will choose the CPU to assign to theFast Path port. The chosen CPU is called the preferred CPU for thisFast Path port. The Fast Path port's interrupt I/O completionprocessing and I/O initiation processing will be performed on thispreferred CPU.

In addition to selecting the preferred CPU, the User Preferred CPU willbe set for this port. Setting the User Preferred CPU prevents the portfrom being reassigned to another CPU unless the User Preferred CPU isbeing stopped. The qualifier can be negated. When the /NOPREFERRED_CPUSqualifier is specified, the User Preferred CPU will be cleared for theport, but it still remains a Fast Path port, and the current preferredCPU will not be changed.

If both /PREFERRED_CPUS and /NOPREFERRED_CPUS are specified on the samecommand line, /NOPREFERRED_CPUS is ignored.

$QIO IO$_SETPRFPATH ! IO$M_PREFERRED_CPU [!IO$M_SYS_ASSIGNABLE]

You can change the assignment of a Fast Path port to a CPU by issuing a$QIO IO$_SETPRFPATH (Set Preferred Path) to the port device, forexample, PNA0. The IO$M_PREFERRED_CPU modifier must be set, and the$QIO argument P1 must be set to either zero or the address of a 32-bitCPU bit mask with a bit set indicating the new preferred CPU. On returnfrom the I/O, the port and its associated devices are all assigned to anew preferred CPU. Note that explicitly setting the preferred CPUoverrides any default assignment of Fast Path ports to CPUs. Thisinterface allows you the flexibility to load balance I/O activity overmultiple CPUs in an SMP system. This is important because I/O activitycan change over the course of a day or week.

The $QIO passes in either a set containing one or more candidate CPUs,or zero as a wildcard value indicating the set of usable CPUs. If thecandidate set contains only one CPU, you are explicitly designating thenew preferred CPU. If the candidate set contains multiple CPUs, you arerequesting use of the automatic preferred CPU assignment algorithm toselect a suitable CPU from the candidate set.

Including the IO$M_SYS_ASSIGNABLE modifier inhibits setting theselected CPU as the device's User Preferred CPU.

The $QIO or the SET DEVICE/PREFERRED_CPU command will make a besteffort to assign the port to a CPU. However, it is possible for thisrequest to return failure for the following reasons:

• There is no intersection between the candidate set and the node's set of usable CPUs.
• There is resource contention. If after a reasonable effort the request is unable to acquire a key system resource, the request will fail. Some key resources include Fast Path spinlock, the CPU mutex, and a CPU transition lock.

If the $QIO or SET DEVICE/PREFERRED_CPU returns failure, you shouldconsider retrying either immediately or after a short delay. It ispossible that a large number of ports were being reassigned, and therequest failed due to resource contention.

$IO_FASTPATH

The $IO_FASTPATH system service performs operations on the set of FastPath devices and CPUs enabled for Fast Path use. The $IO_FASTPATHWsystem service completes synchronously. That is, it returns after theoperation is complete.

The FP$K_BALANCE_PORTS function code specifies that the system serviceis to distribute the set of system assignable Fast Path ports acrossthe intersection of a caller-supplied set of candidate CPUs.

Fast Path restrictions include the following:

• Only high-volume I/Os are optimized.
  Fast Path streamlines the operation of high-volume I/O. I/O that does not meet the definition of high-volume is not optimized. A high-volume Fast Path I/O is characterized as follows:
  1. A virtual, logical, or physical read or write I/O without special I/O modifiers.
  2. An I/O request that is less than 64K bytes in size.
  3. An I/O issued when all I/O resources exist that are needed to perform the I/O.
• Send credits resource must be managed for DSA controllers.
  Applications seeking maximum performance must ensure the availability of sufficient I/O resources.
  The only I/O resource that a Fast Path user needs to be concerned about is send credits. Send credits are extended by DSA controllers to host systems and represent the maximum number of I/Os that can be outstanding at any given point in time. If an application sends an unlimited number of simultaneous I/Os to a controller, it is likely that some I/O will back up waiting for send credits. You can tell whether the send-credit limit is being exceeded by using the DCL command SHOW CLUSTER/CONTINUOUS, followed by an ADD CONNECTIONS, CR_WAIT command. Rapidly increasing credit-wait counts for the disk-class driver connections (a LOC_PROC_NAME name of VMS$DISK_CL_DRVR) is a sign that an application may be incurring send-credit waits.
  To ensure sufficient send credits, some controllers, like the HSC and HSJ, allow the number of send credits to vary. However, not all controllers have this flexibility, and different controllers have different send-credit limits. The best workaround is to know your application access patterns and look for send credit waits. If the number of send credits is being exhausted on one node, then add another controller to spread the load over multiple controllers. An alternative is to rework the application to load balance controller activity throughout the cluster, spreading a given controller's disk load over multiple nodes and allowing an application to exceed the send credits allotted to one node.

This appendix lists the function codes and function modifiers definedin the $IODEF macro. The arguments for these functions are also listed.

A.1 ACP-QIO Interface Driver

This section lists the function codes and function modifiers for theACP-QIO interface driver.

This section lists the function codes and function modifiers for thedisk drivers.

This section lists the function codes and function modifiers for themagnetic tape drivers.