AMS2000

MICROCODE VERSION 0883/B

RELEASED 12/10/2009

Newly supported features and functions for Version 0883/B

New features and functions

Contents 1. Temperature monitoring support
2. Failure factor indication of DP optimization is supported in HDP

Improvements for Version 0883/B

1 Improvements
Contents Improvement of failure search accuracy of Expander hardware

The change of contents for Version 0883/B

1 After CHK1, CTL blockade or power outage

Phenomena After CHK1, CTL blockade or power outage during the DP-Pool format, host I/O to the
pool cannot be executed. (LA error will occur)
Severity High
Conditions of Occurrence The problem may occur when all following conditions are met.
(1) Dynamic Provisioning P.P. is enabled
(2) Pool Format of chunk or page
(3) During (2), CHK1, CTL blockade or power outage occurs
(4) After (3), host I/O is processed
Cause When CHK1/CTL blockade/power outage, the information of the formatting is cleared
falsely The check of the start of pool format has a bug and the dirty data corresponding to
the poor is discarded falsely. As a result, some areas in the pool are not formatted.
Afterward, host I/O to the pool cannot execute as the data inconsistency. (LA error will
occur.)
Affected Products/Version DF800H/M/S, V7.0/A(0870/A) or later
Fixed Product/Version Version: 0872/G, 0883/A or later
Category Dynamic Provisioning
Changed Part MICRO-Disk

2 Subsystem down occurs when a lot of LU is formatted at a time.

Phenomena Subsystem down occurs by "H90320 Watch-dog time-out" when a lot of LU are formatted
at a time.
Severity High
Conditions of Occurrence The problem may occur when all following conditions are met.
1. A lot of LU are formatted at a time
2. (LU number of #1) * (HDD number of RAID group) are bigger than 60240
Cause Judgment process of Drive LED is repeated a lot when a lot of LU is formatted.
Total time of this judgment process exceeds timeout (10 seconds).
Affected Products/Version DF800H/M/S, First version or later
Fixed Product/Version Version: 0883/A or later
Category Common
HDS Confidential 1 of 4
 Changed Part MICRO-Disk

3 Subsystem down when controller blockade or cache non-volatilization

Phenomena Subsystem down occurs when the controller blockade or cache data non-volatilization after
power off occurs on I/O access to virtual LU or pool rebalance in HDP environment.
Severity Medium
Conditions of Occurrence The problem may occur when each following conditions are occurred.
(1)Access to management information about HDP pool is executed. (I/O, pool rebalance, or
pool format, etc.)
(2)controller blockade or cache data non-volatilization after power off occurs
Cause There is a bug about page queue changing sequence. Because recovery process from
inconsistency status is insufficient, pool page status is inconsistent with queue status.
Affected Products/Version DF800H/M/S, V7.0/A or later
Fixed Product/Version Version: 0883/A or later
Category Dynamic Provisioning
Changed Part MICRO-Disk

4 CTL is blockaded with Micro program error[ISM] on ISCSI login

Phenomena CTL is blockaded with Micro program error[ISM] on ISCSI login
Severity Medium
Conditions of Occurrence The problem may occur when all following conditions are met.
(1) ISCSI I/F
(2) Duplicate ISCSI Logins issued from the same host back to back (If this failure occurs
once per 24 hours, the CTL blockade will recover automatically by rebooting. But if this
failure occurs more than once per 24 hours, CTL blockade will not recover automatically.)
Cause Just after receiving an ISCSI Login command from a host, and if the duplicate ISCSI Login
from the same host is received, the process for the 1st login should be aborted for 10ms.
However, there is a bug in the abort process and that cause Micro program error[ISM].
Affected Products/Version DF800H/M/S, SA8x0, First Version or later
Fixed Product/Version Version: 0872CEA, x872/G, x883/A or later
Category ISCSI
Changed Part MICRO-Disk

5 ISCSI port is not available as the port is not link up.

Phenomena ISCSI port is not available as the port is not link up.
Severity Medium
Conditions of Occurrence The problem may occur when all following conditions are met.
(1) ISCSI I/F
(2) An error occurs in an ISCSI protocol chip.
(3) A CTL detect the error and execute reset process to the ISCSI I/F and another CTL also
detect the error and executer reset process to the ISCSI I/F at the same time.
Cause Depending on the timing, a reset to the ISCSI I/F from a CTL and another reset to the
ISCSI I/F, the reset is not executed actually. Then the recovery from reset is not done and
as a result, the link goes down.
Affected Products/Version DF800H/M/S, SA8x0, First version or later
Fixed Product/Version Version: x872/G, x883/A or later
Category ISCSI
Changed Part MICRO-Disk

HDS Confidential 2 of 4
 6 After subsystem power on, a CTL blockaded
Phenomena When power on the subsystem, a CTL is blockaded with a message ”HJ13A3 Micro
program error[DSC]”
Severity Medium
Conditions of Occurrence The problem may occur when ONE of following conditions is met.
1.　PS-ON, ENC firmware download or non-stop firmware update
2. Adding trays or removing HDU/ENC/CTL when the subsystem is ready
3.　Failure on ENC/HDU, CHK1, CTL blockade, backend diagnosis
Cause When the reset process to HDU or ENC is executed, the end processing of the reset has
table search logic (of commands) with a bug. Then the firmware processes the non-
existing command in the table and as a result, the error occurs.
Affected Products/Version DF800H/M/S, SA8x0, First version or later
Fixed Product/Version Version: x883/A or later
Category Common
Changed Part MICRO-Disk

7 CPU utilization rate becomes too high (more than 90%)

Phenomena CPU utilization rate becomes too high (more than 90%) during formatting the DPRG which
consists of SATA HDD.
Severity Medium
Conditions of Occurrence The problem may occur when all following conditions are met.
1. There are too many DPVOL in the Array. (more than 2000 DPVOL)
2. Executing more than 5 DP Pools formatting at the same time.
Cause The firmware executes some unnecessary checks in duplicated Host I/O transaction, so the
CPU utilization rate becomes too high with relating to total LU numbers. (This problem
was detected because generally, DPVOL may be created more than the Normal-LU.)
Affected Products/Version DF800H/M/S, V7.0/A or later
Fixed Product/Version Version: 0883/A or later
Category Dynamic Provisioning
Changed Part MICRO-Disk

8 Controller blockade by "BYTE-ENABLE error"

Phenomena When the HDP function is used, the controller might be blockaded according to the
following WEB message. [DCTL]The BYTE-ENABLE error of the slave command bridge
was detected(CTL-x)　x:Controller number
Severity Medium
Conditions of Occurrence The problem may occur when all following conditions are met.
(1) HDP function use
(2) When the following virtual LU accesses compete in the disk array
- Host access 1 "Read/Write IO" access to area where 32MB boundary was included
- Host access 2 "Read/Write IO" access to area where 32MB boundary is not included
within the range of Host access 1
Cause In the resource release (stripe row release) processing, the release resource might not be
able to be allocated in the resource waiting processing. As a result, the resource waiting
time-out was generated and the illegal table reference happened.
Affected Products/Version DF800H/M/S, V7.0/A(0870/A) or later
Fixed Product/Version Version: 0883/B or later
Category Common
Changed Part MICRO-Disk

9 Controller blockade by "Micro program error (CMD)"

Phenomena When the disk array becomes Queue full, the controller might be blockaded with Micro
HDS Confidential 3 of 4
 program error (CMD).
Severity Medium
Conditions of Occurrence The problem may occur when all following conditions are met.
1. DF800H
2. Host I/F is Fibre
3. Heavy host IO (Queue full response from the disk array)
4. Link error
Cause When the cross call processing cannot be done for heavy host IO, the search of queuing
information in Tachyon was infinitely repeated.
Affected Products/Version DF800H, Version: 0840/A or later
Fixed Product/Version Version: 0883/B or later
Category Frontend (Fibre I/F)
Changed Part MICRO-Disk

10 Subsystem boot failure because both controllers are blockaded.

Phenomena Subsystem can not be booted up because both controllers are blockaded. Or I/O operation
can not be executed because PIN is over when firmware exchange is executed or controller
replacement (insertion) is executed in case system has a SATA HDD as system HDD.
Severity Medium
Conditions of Occurrence The problem may occur when all following conditions are met.
[Not booted up]
(1) Both device discover process (discover) on one backend path and request for device
discover process on another backend path occur.
[I/O operation cannot be executed.]
(1) The system has a SATA HDD as system HDD.
(2) Both device discover process (discover) on one backend path and request for device
discover process on another backend path occur.
Cause Exclusive process about common flag doesn't be executed between controllers. So device
discover process doesn't complete for illegal update.
Affected Products/Version DF800H/M/S, SA800/810 First version or later
Fixed Product/Version Version: 0883/B or later
Category Backend (SAS I/F)
Changed Part MICRO-Disk

HDS Confidential 4 of 4