 |
Guidelines for OpenVMS Cluster Configurations
7.6.2 Setting Up FC Disks for Booting and Dumping
You must use the
wwidmgr
command to set up each device that you will use for booting or dumping.
Once a device is set up, the console retains the information it
requires to access the device in nonvolatile memory. You only have to
rerun the
wwidmgr
command if the system configuration changes and the nonvolatile
information is no longer valid.
The console provides a simplified setup command, called
wwidmgr -quickset
. This command can be used in either of the following cases:
- You are setting up just one device.
- All the devices you are setting up are accessed through the same
ports on the HSG.
If neither description applies to your configuration, refer to the
Wwidmgr User's Manual for additional instructions.
Example 7-3 illustrates the
wwidmgr -quickset
command. Note the following:
- The command
wwidmgr -quickset -udid 10
sets up the FC disk whose HSG80 device identifier is equal to 10.
- The console device names are path dependent. Each path used to
access an FC disk has a different name. In this example, the
wwidmgr -quickset
command establishes four console device names corresponding to the four
paths from the host to the FC disk:
- dga10.1001.0.2.1
- dga10.1002.0.2.1
- dgb10.1003.0.3.1
- dgb10.1004.0.3.1
- The second command,
wwidmgr -quickset -udid 50
, sets up the FC disk whose HSG80 identifier is equal to 50.
- The changes made by the
wwidmgr
command do not take effect until after the next system initialization,
so the next step is to issue an
initialize
command.
- After the initialization, the console
show device
command displays each FC adapter, followed by the paths through that
adapter to each of the defined FC disks. The path-independent OpenVMS
device name for each FC disk is displayed in the second column.
Example 7-3 Using widmgr
-quickset |
P00>>>wwidmgr -quickset -udid 10
Disk assignment and reachability after next initialization:
6000-1fe1-0000-0d10-0009-8090-0677-0016
via adapter: via fc nport: connected:
dga10.1001.0.2.1 kgpsaa0.0.0.2.1 5000-1fe1-0000-0d14 Yes
dga10.1002.0.2.1 kgpsaa0.0.0.2.1 5000-1fe1-0000-0d11 No
dgb10.1003.0.3.1 kgpsab0.0.0.3.1 5000-1fe1-0000-0d13 Yes
dgb10.1004.0.3.1 kgpsab0.0.0.3.1 5000-1fe1-0000-0d12 No
P00>>>wwidmgr -quickset -udid 50
Disk assignment and reachability after next initialization:
6000-1fe1-0000-0d10-0009-8090-0677-0016
via adapter: via fc nport: connected:
dga10.1001.0.2.1 kgpsaa0.0.0.2.1 5000-1fe1-0000-0d14 Yes
dga10.1002.0.2.1 kgpsaa0.0.0.2.1 5000-1fe1-0000-0d11 No
dgb10.1003.0.3.1 kgpsab0.0.0.3.1 5000-1fe1-0000-0d13 Yes
dgb10.1004.0.3.1 kgpsab0.0.0.3.1 5000-1fe1-0000-0d12 No
6000-1fe1-0000-0d10-0009-8090-0677-0026
via adapter: via fc nport: connected:
dga50.1001.0.2.1 kgpsaa0.0.0.2.1 5000-1fe1-0000-0d14 Yes
dga50.1002.0.2.1 kgpsaa0.0.0.2.1 5000-1fe1-0000-0d11 No
dgb50.1003.0.3.1 kgpsab0.0.0.3.1 5000-1fe1-0000-0d13 Yes
dgb50.1004.0.3.1 kgpsab0.0.0.3.1 5000-1fe1-0000-0d12 No
P00>>>initialize
Initializing...
P00>>>show device
polling ncr0 (NCR 53C810) slot 1, bus 0 PCI, hose 1 SCSI Bus ID 7
dka500.5.0.1.1 DKA500 RRD45 1645
polling kgpsa0 (KGPSA-B) slot 2, bus 0 PCI, hose 1
kgpsaa0.0.0.2.1 PGA0 WWN 1000-0000-c920-a7db
dga10.1001.0.2.1 $1$DGA10 HSG80 R024
dga50.1001.0.2.1 $1$DGA50 HSG80 R024
dga10.1002.0.2.1 $1$DGA10 HSG80 R024
dga50.1002.0.2.1 $1$DGA50 HSG80 R024
polling kgpsa1 (KGPSA-B) slot 3, bus 0 PCI, hose 1
kgpsab0.0.0.3.1 PGB0 WWN 1000-0000-c920-a694
dgb10.1003.0.3.1 $1$DGA10 HSG80 R024
dgb50.1003.0.3.1 $1$DGA50 HSG80 R024
dgb10.1004.0.3.1 $1$DGA10 HSG80 R024
dgb50.1004.0.3.1 $1$DGA50 HSG80 R024
polling isp0 (QLogic ISP1020) slot 4, bus 0 PCI, hose 1 SCSI Bus ID 15
dkb0.0.0.4.1 DKB0 RZ1CB-CS 0844
dkb100.1.0.4.1 DKB100 RZ1CB-CS 0844
polling floppy0 (FLOPPY) PCEB - XBUS hose 0
dva0.0.0.1000.0 DVA0 RX23
polling ncr1 (NCR 53C810) slot 4, bus 0 PCI, hose 0 SCSI Bus ID 7
dkc0.0.0.4.0 DKC0 RZ29B 0007
polling tulip0 (DECchip 21040-AA) slot 3, bus 0 PCI, hose 0
ewa0.0.0.3.0 00-00-F8-21-09-74 Auto-Sensing
|
Example 7-4 shows a boot sequence from an FC system disk. Note the
following:
- The boot device is
$1$DGA50
. The user has elected to enter all four paths to the device in the
bootdef_dev
string. This ensures that the system will be able to boot even if a
path has failed.
- The first path on the boot command string,
dga50.1002.0.2.1
, is not currently connected (that is, the disk is not on line to the
HSG80 on that path). The console indicates this fact, retries a few
times, then moves on to the next path in the bootdef_dev string. This
path is currently connected, and the boot succeeds.
- After booting, the OpenVMS SHOW DEVICE command confirms that
OpenVMS has configured all five of the FC devices that were displayed
by the
wwidmgr -show wwid
command, not only the two FC disks that were set up using the console
wwidmgr -quickset
command. The OpenVMS SHOW DEV/MULTIPATH command confirms that OpenVMS
has configured all four paths to each disk.
Example 7-4 Boot Sequence from an FC System
Disk |
P00>>>set bootdef_dev dga50.1002.0.2.1,dga50.1001.0.2.1,dgb50.1003.0.3.1,
dgb50.1004.0.3.1
P00>>>b
(boot dga50.1002.0.2.1 -flags 0,0)
dga50.1002.0.2.1 is not connected
dga50.1002.0.2.1 is not connected
dga50.1002.0.2.1 is not connected
dga50.1002.0.2.1 is not connected
failed to open dga50.1002.0.2.1
(boot dga50.1001.0.2.1 -flags 0,0)
block 0 of dga50.1001.0.2.1 is a valid boot block
reading 919 blocks from dga50.1001.0.2.1
bootstrap code read in
Building FRU table
base = 200000, image_start = 0, image_bytes = 72e00
initializing HWRPB at 2000
initializing page table at 1f2000
initializing machine state
setting affinity to the primary CPU
jumping to bootstrap code
OpenVMS (TM) Alpha Operating System, Version V7.2
...
$ SHOW DEVICE
Device Device Error Volume Free Trans Mnt
Name Status Count Label Blocks Count Cnt
$1$DGA10: (FCNOD1) Online 0
$1$DGA50: (FCNOD1) Mounted 0 V72_SSB 4734189 303 1
$1$DGA51: (FCNOD1) Online 0
$1$DGA60: (FCNOD1) Online 0
$1$DGA61: (FCNOD1) Online 0
$ SHOW LOGICAL SYS$SYSDEVICE
"SYS$SYSDEVICE" = "$1$DGA50:" (LNM$SYSTEM_TABLE)
$ SHO DEV/MULTI
Device Device Error Current
Name Status Count Paths path
$1$DGA10: (FCNOD1) Online 0 4/ 4 PGB0.5000-1FE1-0000-0D11
$1$DGA50: (FCNOD1) Mounted 0 4/ 4 PGA0.5000-1FE1-0000-0D12
$1$DGA51: (FCNOD1) Online 0 4/ 4 PGA0.5000-1FE1-0000-0D13
$1$DGA60: (FCNOD1) Online 0 4/ 4 PGB0.5000-1FE1-0000-0D14
$1$DGA61: (FCNOD1) Online 0 4/ 4 PGB0.5000-1FE1-0000-0D11
Device Device Error Current
Name Status Count Paths path
$1$GGA42: Online 0 4/ 4 PGB0.5000-1FE1-0000-0D11
|
7.7 Setting Up an HSG for Use with OpenVMS
This section describes the basic requirements for setting up an HSG for
use with OpenVMS. Note that the steps for setting up an HSG60 and an
HSG80 are the same, although the examples show the output for an HSG80.
Refer to the HSG80 Array Controller ACS Version 8.4 Configuration
and CLI Reference Guide for more information.
These are the steps you must perform to configure an HSG for OpenVMS:
- If the HSG has dual controllers, set them to multibus failover mode
( Section 7.7.1).
- Set the controller's SCSI version to SCSI-3 ( Section 7.7.2).
Note
You do not need to set the controller's allocation class value. This
parameter is not used by OpenVMS with the HSG.
|
- If you want to access the HSG Command Console LUN (CCL) from
OpenVMS, assign it a device identifier ( Section 7.7.3).
This will
cause OpenVMS to configure the CCL as a generic SCSI device.
Note
In current versions of OpenVMS, you must assign a device identifier to
the CCL. This requirement will be removed in a future update.
|
- Set the parameters for the HSG FC ports ( Section 7.7.4).
Each
port must have its FC profile set to PLDA and its topology set to
FABRIC.
- Initialize and add logical units to the HSG configuration
( Section 7.7.5).
In this step, you must assign a logical unit
number and a device identifier to each logical unit.
Note
You do not need to set the logical unit's preferred path parameter.
OpenVMS does not use this value in its path selection.
|
- Configure the HSG host connection table ( Section 7.7.6).
7.7.1 Setting a Dual HSG to Multibus Failover Mode
To set a dual HSG to multibus failover mode, you can use the following
commands at the HSG console:
- An initial SHOW command to show whether the dual controller is
configured for failover.
- The SET command that puts the controllers into multibus mode.
Notice that this causes the other controller to restart.
- A second SHOW command to show that the dual controller is
configured for multibus failover.
The commands and output are shown in Example 7-5.
Example 7-5 Setting a Dual HSG to Multibus
Failover Mode |
HSG> SHOW THIS_CONTROLLER
Controller:
HSG80 ZG80900677 Software R024F-0, Hardware E01
NODE_ID = 5000-1FE1-0000-0D10
ALLOCATION_CLASS = 0
SCSI_VERSION = SCSI-2
Not configured for dual-redundancy
Controller misconfigured -- other controller present
Device Port SCSI address 6
Time: NOT SET
Command Console LUN is disabled
Host PORT_1:
Reported PORT_ID = 5000-1FE1-0000-0D11
PORT_1_PROFILE = SCC
PORT_1_TOPOLOGY = OFFLINE (offline)
Host PORT_2:
Reported PORT_ID = 5000-1FE1-0000-0D12
PORT_2_PROFILE = SCC
PORT_2_TOPOLOGY = OFFLINE (offline)
NOREMOTE_COPY
Cache:
128 megabyte write cache, version 0012
Cache is GOOD
No unflushed data in cache
CACHE_FLUSH_TIMER = DEFAULT (10 seconds)
Mirrored Cache:
Not enabled
Battery:
FULLY CHARGED
Expires: 27-APR-2000
NOCACHE_UPS
%CER--HSG> --13-JAN-1946 04:32:49 (time not set)-- Controllers misconfigured. -
Type SHOW THIS_CONTROLLER
HSG> SET MULTIBUS_FAILOVER COPY = THIS_CONTROLLER
%EVL--HSG> --13-JAN-1946 04:38:37 (time not set)-- Instance Code: 07050064 (not
yet reported to host)
Template: 5.(05)
Power On Time: 0. Years, 284. Days, 5. Hours, 18. Minutes, 27. Seconds
Controller Model: HSG80
Serial Number: ZG80900677 Hardware Version: E01(29)
Software Version: R024F(FF)
Informational Report
Instance Code: 07050064
Last Failure Code: 08080000 (No Last Failure Parameters)
%EVL--HSG> --13-JAN-1946 04:38:37 (time not set)-- Instance Code: 43010064 (not
yet reported to host)
Template: 4.(04)
Power On Time: 0. Years, 284. Days, 5. Hours, 18. Minutes, 28. Seconds
Controller Model: HSG80
Serial Number: ZG80900677 Hardware Version: E01(29)
Software Version: R024F(FF)
Informational Report
Other Controller Serial Number: ZG80900641
Failed Controller Target Number: 0.(00)
LUNs Taken By This Controller:
00000000
00000000
00000000
00000000
00000000
00000000
00000000
00000000
Instance Code: 43010064
...
%CER--HSG> --13-JAN-1946 04:38:37 (time not set)-- Received LAST GASP message-
from other controller, Last Failure Code: 08080000
HSG>
...
%CER--HSG> --13-JAN-1946 04:39:23 (time not set)-- Other controller restarted
HSG> SHOW THIS_CONTROLLER
Controller:
HSG80 ZG80900677 Software R024F-0, Hardware E01
NODE_ID = 5000-1FE1-0000-0D10
ALLOCATION_CLASS = 0
SCSI_VERSION = SCSI-2
Configured for MULTIBUS_FAILOVER with ZG80900641
In dual-redundant configuration
Device Port SCSI address 6
Time: NOT SET
Command Console LUN is disabled
Host PORT_1:
Reported PORT_ID = 5000-1FE1-0000-0D11
PORT_1_PROFILE = SCC
PORT_1_TOPOLOGY = OFFLINE (offline)
Host PORT_2:
Reported PORT_ID = 5000-1FE1-0000-0D12
PORT_2_PROFILE = SCC
PORT_2_TOPOLOGY = OFFLINE (offline)
NOREMOTE_COPY
Cache:
128 megabyte write cache, version 0012
Cache is GOOD
No unflushed data in cache
CACHE_FLUSH_TIMER = DEFAULT (10 seconds)
Mirrored Cache:
Not enabled
Battery:
FULLY CHARGED
Expires: 27-APR-2000
NOCACHE_UPS
Other controller restarted
|
7.7.2 Setting an HSG to SCSI-3 Mode
Example 7-6 shows how to set the HSG to SCSI-3 mode. Note that the
other controller in a dual configuration is set automatically to the
same mode. You must restart the controllers for this operation to take
effect.
Example 7-6 Setting an HSG to SCSI-3 Mode |
HSG> SET THIS_CONTROLLER SCSI_VERSION = SCSI-3
Warning 4030: Any units that would appear as unit 0 to a host will not be
available when in SCSI-3 mode
Warning 4020: A restart of both this and the other controller is required
before all the parameters modified will take effect
%CER--HSG> --13-JAN-1946 04:42:28 (time not set)-- Restart of this controller-
required
%CER--HSG> --13-JAN-1946 04:42:28 (time not set)-- Restart of the other-
controller required
Restart of this controller required
Restart of the other controller required
|
7.7.3 Setting the Device Identifier for the CCL
If you choose to have OpenVMS configure the CCL, then you must assign
an identifier to the controller pair. This identifier must be unique
for all HSG controller pairs in the cluster.
The following command sets the device identifier for the CCL of the
controller pair to 42. This will cause OpenVMS to configure the CCL as
a generic SCSI device. The device name will be $1$GGA42. Use the SHOW
THIS_CONTROLLER command to verify the setting.
HSG> SET THIS_CONTROLLER IDENTIFIER = 42
|
Note
In current versions of OpenVMS, you must assign a device identifier to
the CCL. This requirement will be removed in a future update.
|
7.7.4 Setting the HSG Port Parameters
To set each HSG port's profile to PLDA and its topology to FABRIC, use
the commands shown in Example 7-7. You can verify the settings with
the SHOW THIS and SHOW OTHER commands.
Note that each port has a unique 64-bit identifier in the form
5000-1FE1-0000-0D11. These identifiers are displayed in the SHOW THIS
and SHOW OTHER displays, next to the label
Reported PORT_ID
. These identifiers are used by OpenVMS to denote the multiple paths
that may exist between the host and the logical unit.
Note
The command
HSG> SET THIS_CONTROLLER PORT_1_PROFILE=PLDA
, shown in Example 7-7, is required for ACS 8.4. If you are using ACS
8.5 or above, omit this step.)
|
Example 7-7 Setting the HSG Port
Parameters |
HSG> SET THIS_CONTROLLER PORT_1_PROFILE = PLDA
Warning 4000: A restart of this controller is required before all the
parameters modified will take effect
%CER--HSG> --17-MAR-1999 12:24:09-- Restart of this controller required
Restart of this controller required
HSG> SET THIS_CONTROLLER PORT_1_TOPOLOGY = FABRIC
Warning 4000: A restart of this controller is required before all the
parameters modified will take effect
%CER--HSG> --17-MAR-1999 12:25:02-- Restart of this controller required
Restart of this controller required
HSG> SET THIS_CONTROLLER PORT_2_PROFILE = PLDA
Warning 4000: A restart of this controller is required before all the
parameters modified will take effect
%CER--HSG> --17-MAR-1999 12:25:26-- Restart of this controller required
Restart of this controller required
HSG> SET THIS_CONTROLLER PORT_2_TOPOLOGY = FABRIC
Warning 4000: A restart of this controller is required before all the
parameters modified will take effect
%CER--HSG> --17-MAR-1999 12:25:39-- Restart of this controller required
Restart of this controller required
HSG> SHOW THIS_CONTROLLER
Controller:
HSG80 ZG80900677 Software R024F-0, Hardware E01
NODE_ID = 5000-1FE1-0000-0D10
ALLOCATION_CLASS = 0
SCSI_VERSION = SCSI-3
Configured for MULTIBUS_FAILOVER with ZG80900641
In dual-redundant configuration
Device Port SCSI address 6
Time: 17-MAR-1999 12:25:55
Command Console LUN is lun 0 (IDENTIFIER = 42)
Host PORT_1:
Reported PORT_ID = 5000-1FE1-0000-0D11
PORT_1_PROFILE = PLDA
PORT_1_TOPOLOGY = FABRIC (offline)
Host PORT_2:
Reported PORT_ID = 5000-1FE1-0000-0D12
PORT_2_PROFILE = PLDA
PORT_2_TOPOLOGY = FABRIC (offline)
NOREMOTE_COPY
Cache:
128 megabyte write cache, version 0012
Cache is GOOD
No unflushed data in cache
CACHE_FLUSH_TIMER = DEFAULT (10 seconds)
Mirrored Cache:
Not enabled
Battery:
FULLY CHARGED
Expires: 27-APR-2000
NOCACHE_UPS
HSG> SET OTHER_CONTROLLER PORT_1_PROFILE = PLDA
Warning 4010: A restart of the other controller is required before all the
parameters modified will take effect
%CER--HSG> --17-MAR-1999 12:26:46-- Restart of the other controller required
Restart of the other controller required
HSG> SET OTHER_CONTROLLER PORT_1_TOPOLOGY = FABRIC
Warning 4010: A restart of the other controller is required before all the
parameters modified will take effect
%CER--HSG> --17-MAR-1999 12:26:59-- Restart of the other controller required
Restart of the other controller required
HSG> SET OTHER_CONTROLLER PORT_2_PROFILE = PLDA
Warning 4010: A restart of the other controller is required before all the
parameters modified will take effect
%CER--HSG> --17-MAR-1999 12:27:08-- Restart of the other controller required
Restart of the other controller required
HSG> SET OTHER_CONTROLLER PORT_2_TOPOLOGY = FABRIC
Warning 4010: A restart of the other controller is required before all the
parameters modified will take effect
%CER--HSG> --17-MAR-1999 12:27:15-- Restart of the other controller required
Restart of the other controller required
HSG> SHOW OTHER_CONTROLLER
Controller:
HSG80 ZG80900641 Software R024F-0, Hardware E01
NODE_ID = 5000-1FE1-0000-0D10
ALLOCATION_CLASS = 0
SCSI_VERSION = SCSI-3
Configured for MULTIBUS_FAILOVER with ZG80900677
In dual-redundant configuration
Device Port SCSI address 7
Time: 17-MAR-1999 12:27:31
Command Console LUN is lun 0 (IDENTIFIER = 42)
Host PORT_1:
Reported PORT_ID = 5000-1FE1-0000-0D13
PORT_1_PROFILE = PLDA
PORT_1_TOPOLOGY = FABRIC (offline)
Host PORT_2:
Reported PORT_ID = 5000-1FE1-0000-0D14
PORT_2_PROFILE = PLDA
PORT_2_TOPOLOGY = FABRIC (offline)
NOREMOTE_COPY
Cache:
128 megabyte write cache, version 0012
Cache is GOOD
No unflushed data in cache
CACHE_FLUSH_TIMER = DEFAULT (10 seconds)
Mirrored Cache:
Not enabled
Battery:
FULLY CHARGED
Expires: 27-APR-2000
NOCACHE_UPS
|
7.7.5 Setting Up Logical Units on the HSG
Setting up logical units on the HSG consists of the following steps, as
shown in Example 7-8:
- Configure all the physical disk devices on the HSG with the CONFIG
command.
- Configure each disk device as a logical unit as follows:
- Use the INITIALIZE command to initialize the disk.
- Use the ADD command to add the initialized disk as a host-visible
logical unit.
Choose a logical unit number that is unique for the
HSG storage subsystem. The value chosen in Example 7-8 is 20, which
results in a logical unit name of D20.
- Use the SET command to set a device identifier that will serve as
the basis for the OpenVMS device name.
This device identifier must
be unique for all FC disks in the OpenVMS Cluster. The value chosen in
Example 7-8 is 32020, which results in an OpenVMS device name of
$1$DGA32020 for the logical unit. Note: This device identifier is
limited to 9999, if the FC device is MSCP served. This restriction will
be removed in a future release.
Example 7-8 Setting Up Logical Units on the
HSG |
HSG> RUN CONFIG
Config Local Program Invoked
Config is building its tables and determining what devices exist
on the subsystem. Please be patient.
add disk DISK10000 1 0 0
add disk DISK10100 1 1 0
add disk DISK10200 1 2 0
add disk DISK20000 2 0 0
Config - Normal Termination
HSG> SHOW DISK FULL
Name Type Port Targ Lun Used by
------------------------------------------------------------------------------
...
DISK20000 disk 2 0 0
DEC RZ1CD-CS (C) DEC 0306
Switches:
NOTRANSPORTABLE
TRANSFER_RATE_REQUESTED = 20MHZ (synchronous 20.00 MHZ negotiated)
Size: 8378028 blocks
HSG> SHOW UNITS
No LUNs
HSG> INITIALIZE DISK20000
HSG> ADD UNIT D20 DISK20000
HSG> SET D20 IDENTIFIER = 32020
HSG> SHOW UNITS FULL
LUN Uses Used by
------------------------------------------------------------------------------
D20 DISK20000
LUN ID: 6000-1FE1-0000-0D10-0009-8090-0677-0024
IDENTIFIER = 32020
Switches:
RUN NOWRITE_PROTECT READ_CACHE
READAHEAD_CACHE WRITEBACK_CACHE
MAXIMUM_CACHED_TRANSFER_SIZE = 32
Access:
ALL
State:
ONLINE to this controller
Not reserved
NOPREFERRED_PATH
Size: 8378028 blocks
Geometry (C/H/S): ( 3708 / 20 / 113 )
|
|