Display the help screen and do not start the ssmcli process.

One of a set of subcommands that are used with the ssmcli command to configure and manage a storage system, such as bind or getdisk. See “ssmcli Subcommands”.

The target RAID device. If you do not specify the -d switch, the environment variable RaidAgentDevice is the default value for the device. (RaidAgentDevice is explained later in this section.) If RaidAgentDevice is not set and the -d switch is not present, an error is generated on all commands that need device information. Unless the environment variable indicates otherwise, every instance of the ssmcli command must have the -d switch. The -d switch overrides the RaidAgentDevice environment variable. You can get a list of RAID devices present using the getagent command with ssmcli; see “getagent” for a complete description.

The network name of the destination server. The environment variable RaidAgentHost is the default value for the server if the -h switch is not present. The -h switch overrides the RaidAgentHost setting. If RaidAgentHost is not set and the -h switch is not present, the local hostname is used.

Suppresses output except for values. This option is most useful when it is used as part of a script.

Parses the entered command without making a network call to the application programming interface (API). If the string does not parse correctly, an error message prints to stderr; otherwise a message verifying that the string parsed correctly prints to stderr.

Suppresses error messages. This switch can be useful in a script.

Sets the timeout value in seconds. Default is 25 seconds unless the RaidAgentTimeout environment variable is set differently. See “Environment Variables RaidAgentHost, RaidAgentDevice, and RaidAgentTimeout” for more information.

Enables verbose error descriptions. This is the default unless -q is specified.

]
A subcommand and its switches. These subcommands are described in “ssmcli Subcommands”; Table 6-1 on page 144 summarizes the subcommands.

Automatically reassigns ownership of a LUN to an active SP if an SP failure occurs. By default this feature, known as auto assign, is disabled for all RAID types except hot spares, to which it does not apply. The values are

See the description of this parameter on page 147 in “bind.”

Change the default owner of the LUN.

Maximum number of I/Os that can be outstanding to a LUN for the LUN still to be considered idle. Used to determine cache flush start time. Valid values are 0 through 254; the default is 0.

Amount of time in 100-ms units that a LUN must be below the idle-thresh in order to be considered idle. For example, 5 equals 500 ms. Once a LUN is considered idle, any dirty pages in the cache can begin idle time flushing. The default is 20.

Maximum number of I/Os which the LUN can have outstanding and still perform prefetching. The default is 40; valid values are 1 through 100.

Priority that the array allots for checking parity: ASAP (as soon as possible), High, Medium, or Low. If an SP detects parity inconsistencies, it starts a background process to check all the parity sectors in the LUN. The priority you specify determines the amount of resource the SP devotes to verifying instead of to normal I/O activity.

Enables minimal latency reads functionality. A setting of 0 disables; a setting of 1 enables. This parameter applies to RAID 3 only.

Priority that the array assigns for reconstructing the data on either a hot spare or a new disk module that replaces a failed disk module in a LUN: ASAP (as soon as possible; 10 to 15 minutes per GB), High (15 to 30 minutes per GB), Medium (45 to 60 minutes per GB), or Low (75 to 90 minutes per GB). The rebuild priority applies to all RAID LUNs except RAID 0, individual disk LUNs, and hot spares. The default priority is High, which is adequate for most situations.

---

Note: You cannot change the rebuild priority for a LUN while the LUN is rebuilding. Any change to the rebuild priority takes effect as soon as the operation completes.

---

Sets the maximum write request size, in blocks, that bypasses the cache and goes directly to the disks. Write requests equal to or less than the write-aside value are written directory to disk, bypassing write cache.

Sets the default prefetch parameter values.

Determines whether to prefetch data of a variable or constant length or to disable prefetching:

Determines the amount of data prefetched for one host read request. The action of this option depends on the setting for the -pf option:

For variable-length prefetching, sets the absolute maximum number of disk blocks to prefetch in response to a read request. This value is ignored for constant length. The range is 0 through 2048 blocks; the default is 512.

For constant-length prefetching, sets the number of blocks of data to prefetch for each read request. An SP reads one segment at a time from the LUN because smaller prefetch requests interfere less with other host requests.

For variable-length prefetching, determines the amount of data to retrieve from the LUN in a single prefetch operation. Specify a factor equal to or less than the prefetch multiplier. If the segment multiple is equal to the prefetch multiplier, prefetch operations are not divided into segments. The amount of data prefetched for each read request must equal or be less than the prefetch multiplier.

Sets the prefetch disable, which is the size in blocks of the smallest read request for which prefetching is disabled. Prefetching is disabled for all read requests whose size is equal or larger than the Prefetch Disable size. The range is 0 through 65534; the default value is 129 (blocks).

Enables retaining prefetch data. A setting of 0 disables; a setting of 1 (the default) enables.

Cabinet type.

ASCII description of the device.

Size in MB of the SP's memory. Each SP in the storage system must have the same amount of memory to make full use of the memory.

Model number of the device.

Name of the device.

Location in file system of device.

Signature of peer SP.

Revision level of the SP's PROM code. Each SP in the storage system runs the same revision of PROM code. PROM code is updated automatically when an SP's LIC is updated.

Microcode revision number: the revision of the Licensed Internal Code (LIC) that the SP is running. Each SP in the storage system runs the same revision of LIC.

SCSI ID number for the SP host connection, which is determined by switch settings on the SP. For information on these switch settings, see the Origin FibreVault and Fibre Channel RAID Owner's Guide

Enclosure serial number.

Unique SP signature.

Model type (deskside, rackmount).

Agent revision number: the version of the agent that is running.

Day of the week that BBU self-test is performed.

Time of day that BBU self-test is performed. Format: hhmm.

Free memory available in the back-end memory bank.

Free memory available in the control memory bank.

Free memory available in the front-end memory bank.

Write cache high watermark. The percentage of dirty pages that, when reached, causes the SP to begin flushing the cache. The default is 100%. A lower value causes the SP to start flushing the cache sooner. The high watermark cannot be less than the low watermark. Watermark processing is turned off when both the low and high watermark values are 100.

Write cache low watermark. The percentage of cache dirty pages that determines when cache flushing stops. When the write cache low watermark is reached during a flush operation, the SP stops flushing the cache. The default value is 50%. Do not set the low watermark higher than the high watermark. Watermark processing is turned off when both the low and high watermark values are 100.

Write cache mirrored state.

Size of a cache page in KB: 2, 4, 8, or 16 KB. The default size is 2 KB.

Percent dirty pages currently in cache. The percentage of pages that have been modified in the SP's write cache, but that have not yet been written to disk. A high percentage of dirty pages means the cache is handling many write requests.

Physical memory size for the back-end memory bank.

Physical memory size for the front-end memory bank.

Physical memory size for the control memory bank.

Percent pages owned by SP.

Read hit ratio: the percentage of cache read hits for the SP. A read hit occurs when the SP finds a sought page in cache memory, and thus does not need to read the page from disk. The ratio is meaningful only if the SP's read cache is enabled. High hit ratios are desirable because each hit indicates at least one disk access that was not needed.

You may want to compare the read and write hit ratios for the LUN with the read and write hit ratio for the entire storage system. For a LUN to have the best performance, the hit ratios should be higher than those for the storage system. A very low read or write hit rate for a busy LUN may mean that caching is not helping the LUN's performance.

Amount of memory, in MB, reserved for the optimized RAID 3 data buffers on SP A.

The amount of memory, in MB, reserved for the optimized RAID 3 data buffers on SP B.

Current internal state of the read cache on SP A.

Current internal state of the read cache on SP B.

Amount of read cache memory SP A is allowed to use (in MB).

Amount of read cache memory SP B is allowed to use (in MB).

Number of pages owned by SP A. The total number of pages in the SP, each page with the cache page size you selected when setting up storage-system caching. This number equals the cache size divided by the cache page size, minus space for checksum tables. If the storage system has two SPs and both are working, they divide the total number of pages between them. If an SP is idle for a long period or fails, the active SP may increase its share of pages.

Number of pages owned by SP B. The total number of pages in the SP, each page with the cache page size you selected when setting up storage-system caching. This number equals the cache size divided by the cache page size, minus space for checksum tables. If the storage system has two SPs and both are working, they divide the total number of pages between them. If an SP is idle for a long period or fails, the active SP may increase its share of pages.

Current state of the SP's read or write cache. The possible read cache states are Enabled, Disabling, and Disabled. An SP's read cache is automatically enabled on power-on if the cache size is a valid number and the SP has at least 8 MB of memory. The possible write cache states are Enabled or Disabled, and several transition states, such as Initializing, Enabling, Disabling, Dumping, and Frozen.

Unassigned cache pages (that is, owned by neither SP A nor B).

Write hit ratio: the percentage of cache write hits for the SP's write cache. A write hit occurs when the SP finds and modifies data in cache memory, which usually saves a write operation. For example, with a RAID 5 LUN, a write hit eliminates the need to read, modify, and write the data. High hit ratios are desirable because each hit indicates at least one disk access that was not needed.

You may want to compare the read and write hit ratios for the LUN with the read and write hit ratio for the entire storage system. For a LUN to have the best performance, the hit ratios should be higher than those for the storage system. A very low read or write hit rate for a busy LUN may mean that caching is not helping the LUN's performance.

Write cache state for SP A; it is always the same as that of SP B.

Write cache state for SP B; it is always the same as that of SP A.

Write cache size for SP A; it is always equal to that of SP B.

Write cache size for SP B; it is always equal to that of SPA.

Idle and busy ticks for both buses.

Number of times a user request arrived while at least one other request was being processed.

Average number of requests seen by the SP.

Percentage of time the SP is busy.

Controller idle and busy ticks.

SP's system date.

SP's system day.

Number of times the high watermark, low watermark, and idle unit turned on flushing. This parameter also displays the number of requests to flush write cache, and the number of write cache blocks flushed.

Number of hard errors the SP has seen.

Displays internal bus 1 idle and busy ticks.

Displays internal bus 2 idle and busy ticks.

Percentage of time the SP is idle.

Maximum number of requests seen by the SP.

Number of reads received by the SP.

Displays the following:

Indicates whether system cache is disabled or enabled. For SP A, the output gives the state of both read and write cache; for example:

SP Read Cache State Disabled
SP Write Cache State Disabled

For SP B, the output is read cache only, because write cache is always the same as that for SP A.

Indicates whether system fault LED is on or off.

Indicates whether statistics logging is on or off.

SP's system time.

Number of writes received by SP.

Gets the state information on fan module A, which is the fan module (tray, bank) for the drives (rear fan module) in a DPE, or the only fan module in a DAE.

If any fan in a module fails, a fault is displayed for its module, and other fans in the module speed up to compensate. They cool the storage system indefinitely; however, if failure occurs in the normally functioning module, the temperature rises and the storage system shuts down after two minutes.

If a fan bank is down, you can determine whether it has failed by whether its button is amber, the event log lists fan failures, or the replace light on the fan module is lit. If a fan module has failed, replace it as soon as is practical.

Gets the state information on fan module B, which is the SP fan pack in a DPE (front fan). See -fana for further information.

State information on the link controller card (LCC) for SP A.

State information on the LCC for SP B.

Current operational state of SP A. It can be in any of these states:

Gets state information on SP B. See the description of -spa.

Gets state information on SPS (standby power supply) A.

Gets state information on SPS B.

Gets state information on power supply A.

If a chassis has the maximum number of VSCs, it can continue operating if one VSC fails, but shuts down immediately if a second VSC fails. If a VSC is Down, you can determine whether it has failed by seeing if its button is amber, looking at the event log for VSC failures, or seeing if the Replace light on the VSC is lit. If the VSC has failed, replace the VSC as soon as is practical.

Gets the state information on power supply B. You can use getcrus without any switches to get the state information for every CRU except disks.

Table 6-4 summarizes states.

Use the format e_d, where e is the enclosure number (1-10) and d is the disk number (0-9). For example, 2_3 represents disk 3 in enclosure 2. Valid enclosure values are the hexadecimal numbers 0 through 9, A (the hex representation of 10), and B (the hex representation for 11). Valid disk numbers are 0 through 9.

When the disk-position is not entered, getdisk returns an inclusive report on all disks, regardless of what switches are set. To obtain only the information specified by the optional switches, set the disk-position switch in the getdisk command.

Percent of the disk that is bound.

Disk capacity in blocks.

Total hard read errors for all the disk modules in the LUN that persisted through all the retries. A growing number of hard errors may mean that one or more of the LUN's disk modules is nearing the end of its useful life.

Hot spare information (NO if not hot spare).

Total hard write errors for all the disk modules in the LUN that persisted through all the retries. A growing number of hard errors may mean that one or more of the LUN's disk modules is nearing the end of its useful life.

Possible values include: a hexadecimal number, Unbound for an enclosure disk not bound to a LUN, or Empty for an empty enclosure disk slot.

Private space of disk in blocks.

Product ID of the disk.

Percent of the disk that is rebuilt.

Total read requests made to the disk module. You might find the LUN read information in getlun more useful because it is for the entire LUN, and not just for one of the disk modules in the LUN.

Product revision number of the disk module.

Number of read retries. Total number of times read and write requests to the disk module were retried.

Number of remapped sectors. Total number of disk sectors that were faulty when written to, and thus were remapped to a different part of the disk module.

State of disk, as summarized in Table 6-5.

Number of user sectors.

Serial number of the disk.

Bind signature (automatically assigned during bind).

Total soft read errors for all the disk modules in the LUN that disappeared before all the retries. A growing number of soft errors may indicate that one of the LUN's disk modules is nearing the end of its useful life.

Total soft write errors for all the disk modules in the LUN that disappeared before all the retries. A growing number of soft errors may indicate that one of the LUN's disk modules is nearing the end of its useful life.

LUN type of the disk.

Vendor ID of the disk.

Number of write retries. Total number of times write requests to the disk module were retried.

Total write requests made to the disk module. You might find the LUN write information in getlun more useful because it is for the entire LUN, and not just for one of the disk modules in the LUN.

Display the getlog header.

Displays the oldest N entries in the log, with the oldest entry first. If there are fewer than N entries, the entire log is displayed. N must be greater than 0.

Displays the newest N entries in the log, with the oldest entry first. If there are fewer than N entries, the entire log is displayed. N must be greater than 0.

Auto-assign is enabled or disabled.

Auto-trespass is enabled or disabled.

Percent of the LUN that is bound.

Number of blocks prefetched by read cache. This field is not printed if statistics logging is disabled.

Number of blocks read for each CRU. This switch must be used in conjunction with the -disk switch.

Number of host block read and write requests. This field is not printed if statistics logging is disabled.

Percentage of time that the disk is busy. This switch must be used in conjunction with the -disk switch.

Number of blocks written for each CRU. This switch must be used in conjunction with the -disk switch.

LUN capacity in MB.

Names and states of disks in this LUN.

SP is or is not the default owner of the LUN.

Disk statistics for the LUN. Disk statistics switches are only valid when the -disk switch is included.

Stripe size (in blocks), which is the number of sectors that the storage system can read or write to a single disk module in the LUN. The default size is 128 sectors. This size was specified when the LUN was bound. Stripe element size does not apply to a RAID 1 LUN, individual disk, or hot spare.

Disk is enabled or disabled. This switch must be used in conjunction with the -disk switch.

Number of times that a write had to flush a page to make room in the cache. This field is not printed if statistics logging is disabled.

Percentage of time that the LUN is idle. This switch must be used in conjunction with the -disk switch.

Idle delay time for caching (in 100-ms intervals).

Idle threshold value for caching.

For variable length prefetching, gets the absolute maximum number of disk blocks to prefetch in response to a read request.

Gets whether minimal latency reads is or is not supported. This field is not printed if statistics logging is disabled.

Which SP currently owns the LUN, SPA or SPB.

Gets the prefetch disable size, which is the size in blocks of the smallest read request for which prefetching is disabled. Prefetching is disabled for all read requests whose size is equal or larger than the Prefetch Disable size.

Maximum number of I/Os that can be outstanding to the unit and have the unit still perform prefetching.

Percentage of the LUN that is rebuilt.

Type of prefetching that the LUN is using for read caching:

Percentage of read requests that caused a flush.

Prefetch size/multiplier. The amount of data prefetched for one host read request. For constant-length prefetching, gets the read request in blocks.

Prefetch size/multiplier. The amount of data prefetched for one host read request. For variable-length prefetching, gets the multiplier used to calculate the amount of data to prefetch. For example, if the prefetch size is 8, the amount of data to prefetch is 8 times the amount of data requested.

Average queue depth for each CRU. This switch must be used in conjunction with the -disk switch.

Maximum queue depth for each CRU. This switch must be used in conjunction with the -disk switch.

Maximum priority level for rebuilding failed disk.

LUN's read cache state (enabled or disabled).

Read and write cache hits and misses. This field is not printed if statistics logging is disabled.

Number of reads received for each CRU. This switch must be used in conjunction with the -disk switch.

Number of remapped sectors. This switch must be used in conjunction with the -disk switch.

Retention priority for retaining prefetched data when the read cache becomes full. If the priority is Equal Priority, prefetched data has priority over host-requested data. If it is Equal, both types of data have the same priority.

Gets read-hit information if statistics logging is enabled. This field is not printed if statistics logging is disabled.

Read histogram, which is an array of 10 locations that contain the number of reads. Element n of the array contains the number of reads that were larger than or equal to 2n-1 and less than 2n blocks in size. For example, the first bucket contains 1 block reads, the second contains 2-3, the third 4-7 and so on. -rhist also displays read-histogram overflows, which are the number of I/O operations that were larger than 512 blocks. This field is not printed if statistics logging is disabled.

Total number of times read requests to all the disk modules in the LUN were retried. This switch must be used in conjunction with the -disk switch.

Percent of write requests that caused a flush.

Host read and write requests. This field is not printed if statistics logging is disabled.

Average service time for each CRU. This switch must be used in conjunction with the -disk switch.

Read cache configuration.

Prefetch segment size, the size of the segments that make up a prefetch operation. For constant-length prefetching, gets the segment size (in blocks) of data prefetched in one read operation from the LUN.

Prefetch segment multiplier. For variable length prefetching, gets the multiplier which determines the amount of data, relative to the amount of requested data, prefetched in one read operation. For example, if the segment multiplier is 4, the segment size is 4 times the amount of data requested.

Number of times an I/O crossed a stripe boundary on a RAID 5, RAID 0, or RAID 1/0 LUN.

RAID type of the LUN as follows:

Number of prefetched blocks not used by the read cache. This field is not printed if statistics logging is disabled.

Verification priority. This field is not printed if statistics logging is disabled.

Minimum block size request that bypasses cache.

LUN's write cache state (enabled or disabled).

Number of times that a write was completely satisfied by the cache, avoiding a disk operation. This field is not printed if statistics logging is disabled.

Write hit information, if statistics logging is enabled. This field is not printed if statistics logging is disabled.

Write histogram, which is an array of 10 locations that contain the number of writes. Element n of the array contains the number of writes that were larger than or equal to 2n-1 and less than 2n blocks in size. For example, the first bucket contains 1 block writes, the second contains 2 to 3 block writes, the third 4 to 7 block writes, and so on. -whist also displays the write histogram overflows, which are the number of I/O that were larger than 512 blocks. This field is not printed if statistics logging is disabled.

Total number of times write requests to all the disk modules in the LUN were retried. This switch must be used in conjunction with the -disk switch.

Number of writes received for each CRU. This switch must be used in conjunction with the -disk switch.

Displays historical totals of all nonvolatile recovery verifies.

Displays the currently running full-unit verify.

Displays the most recently completed full-unit verify.

Displays historical totals of all full-unit verifies.

Specifies both SPs.

Suppresses the prompt that asks the user for permission.

Write cache high watermark: the percentage of dirty pages, which, when reached, causes the SP to begin flushing the cache. The default is 96%. A lower value causes the SP to start flushing the cache sooner. The high watermark cannot be less than the low watermark. Watermark processing is turned off when both the low and high watermark values are 100.

Write cache low watermark: the percentage of cache dirty pages that determines when cache flushing stops. When the write cache low watermark is reached during a flush operation, the SP stops flushing the cache. The default value is 80%. The low watermark cannot be greater than the high watermark. Watermark processing is turned off when both the low and high watermark values are 100.

Enable or disable mirroring. A setting of 1 enables mirroring (the default). For fibre channel storage, a setting of 0, which disables mirroring, is not supported and an error message is reported.

Size (in KB) of pages into which to partition the cache. Valid sizes are 2 (the default), 4, 8, and 16. Page size is set to 2 regardless of enable/disable state. Generally, cache page size should be 8 KB for general IRIX file server applications, and 2 KB or 4 KB for database applications. This switch does not work if read or write cache is enabled.

---

Note: Setting a RAID 3 size or a write cache size for one SP automatically sets the same value for the other SP, so you need specify a value for only one SP.

---

Optimized RAID 3 size for SP A. For RAID 3 LUNs to be accessible, the minimum value is 2 MB and the maximum value (optimum) is 14 MB.

---

Note: For storage systems other than RAID 3, set this field to 0 only.

---

Sets the optimized RAID 3 size for SP B. For RAID 3 LUNs to be accessible, the minimum value is 2 MB. For more information, see -r3a.

Sets read cache size for SP A.

Sets read cache size for SP B.

Sets the read cache state for SP A. A setting of 1 enables read cache; a setting of 0 disables.

Sets read cache state for SP B. A setting of 1 enables; a setting of 0 disables.

Sets write cache state. A setting of 1 enables write cache; a setting of 0 disables.

Sets write cache size for SP A. Write cache on each SP must be equal. If you set the size for one SP, the software automatically sets the size in the other SP to match. Write caching requires two SPs with equal amounts of memory.

Sets write cache size for SP B; see the explanation of -wsza write-cache-size-spa.

Starts a full-unit background verify.

Clears the background verify reports.

Specifies a sniff rate in units of 100 ms. This setting controls the rate at which sniffs are executed. Valid values are 1 through 254; the default is 30.

Specifies priority for the background verify rate for executing the full background verify. Allowable values are ASAP, High, Medium, and Low; the default is High.

Sets the following:

To set the SPS test time, all three switches must be specified

Performs whatever trespassing is necessary to take all of the LUNs that have this SP as a default owner.

Specifies trespass on one LUN (specified by the lun-number).

Specifies trespass on all LUNs in the system.