EMC VNX FAST VP for VNX OE for Block Release Notes · PDF file03/12/2012 · 1 EMC® VNX Operating ... VNX Block to Unified upgrade This version of VNX OE for Block supports ... EMC

1

EMC® VNX Operating Environment

For Block Version 05.32.000.5.015

Release Notes P/N 300-013-468

REV 04 December 3, 2012

These release notes contain supplemental information about the EMC® VNX® Operating Environment for Block for the following EMC VNX storage systems: VNX5100, VNX5300, VNX5500, VNX5700, and VNX7500. The software version is the same for all platforms: 05.32.000.5.015. Topics include:

Revision history ......................................................................................... 2 Product description ................................................................................... 2 New features and changes ........................................................................ 3 Environment and system requirements.................................................. 4 Fixed problems ........................................................................................... 5 Known problems and limitations .......................................................... 16 Technical notes ......................................................................................... 21 Documentation ......................................................................................... 25 Software media, organization, and files ............................................... 25 Installation ................................................................................................ 25 Troubleshooting and getting help ......................................................... 26

2

Revision history

VNX Operating Environment for Block 05.32.000.5.015 Release Notes

Revision history

The following table presents the revision history of this document.

Revision Date Description 04 December 3, 2012 Updated for service pack 05.32.000.5.015

03 September 28, 2012 Updated for service pack 05.32.000.5.011. Minor editorial changes and reformatting of tables incorporated.

02 August 13, 2012 Updated for service pack 05.32.000.5.008 01 July 16, 2012 Initial release

Product description

EMC VNX Operating Environment for Block controls the operation of the VNX5100, VNX5300, VNX5500, VNX5700, and VNX7500 disk-array storage systems. Each storage system ships with a complete copy of VNX for Block software installed on the first four disks on back-end loop 0. Disks 0_0 and 0_2 store mirrored copies of the software for storage processor A. Disks 0_1 and 0_3 store mirrored copies of the software for storage processor B. When you power up the storage system, each storage processor boots and executes the VNX for Block software.

Note: You must have the VNX Block Enabler package installed to connect hosts to the storage system.

Disk array enclosure (DAE) current firmware revision

This section identifies the revision of enclosure firmware contained in this version. The firmware upgrades are initiated as part of the non-disruptive upgrade to this version. Depending upon the system configuration, the firmware upgrades may continue processing after the upgrade of the other system software has finished.

During the firmware upgrade, note the following:

If a lower revision is installed, the firmware is automatically upgraded to the revision contained in this version.

If a higher revision is running, the firmware is not downgraded to the revision contained in this version.

3

New features and changes


Table 1: Firmware types Enclosure type Current firmware version 15 Drive 3.5” DAE 01.01.3.310: version 7.05-2

25 Drive 2.5” DAE 01.01.3.310: version 7.05-2

60 Drive 3.5” DAE 01.01.3.325: version 1.33-3

New features and changes

The Fixed problems section contains changes made in this version.

New features in 05.32.000.5.015, 05.32.000.5.011 and 05.32.000.5.008

There are no new features in these releases.

New features in 05.32.000.5.006

Dual-port 10GBASE-T Module

The dual-port 10GBASE-T module provides high-performance, low-cost host connectivity for all members of the VNX array family except VNX5100.

The ports employ a RJ45 connector and CAT6/6a/7 copper cabling to connect to10GbE.

VNX7500 48 GB of memory per SP

This version of VNX OE for Block can support 48 GB of memory on a VNX7500 as an optional alternative to the 24GB configuration. An installed VNX7500 with 24 GB of SP memory is field-expandable to 48 GB.

The above upgrade is performed by the service provider.

VNX Block to Unified upgrade

This version of VNX OE for Block supports upgrade by the customer to Unified (Block and File) services. The upgrade includes installing the necessary hardware to enable file services, installing the necessary operating system (VNX OE for File), and configuring file services.

VNX in-family data in-place conversions

This version of VNX OE for Block supports in-family data in place (DIP) conversion from VNX5100/5300/5500 to VNX5700/7500.

4

Environment and system requirements


In-family DIP conversions across the VNX family re-use the installed SAS disk array enclosures (DAEs) with associated drives and I/O modules.

Conversion of VNX5100/5300/5500 to VNX5700/7500 arrays requires replacement of disk processor enclosures (DPEs) with equivalent form factor DAEs (either 25x2.5” drive DAE or 15x3.5” drive DAE), along with the addition of VNX5700/7500 storage processor enclosure (SPE).

The above conversion is performed by the service provider.

Increased stripe element size

This version of VNX OE for Block supports an optional increased stripe element size for FLARE® LUNS, configured using Unisphere®. This increases the read efficiency for large read request sizes, as it reduces the ratio of seek time to data transfer time.

This capability is targeted specifically to environments focused on large random block reads. Write performance will degrade in some circumstances. Applications performing small I/O requests should not utilize this capability.

Environment and system requirements

Table 2 lists the minimum requirements for the proper operation of VNX OE for Block version 05.32 with the VNX5100, VNX5300, VNX5500, VNX5700, and VNX7500 storage systems.

Table 2: VNX for Block requirements Storage system type

Minimum hardware requirements

Minimum software requirements

Supported Disk/Flash Drives

VNX5100 VNX5300 VNX5500 VNX5700 VNX7500

• 4 disks minimum, installed in slots

0 through 3 of enclosure 0

• 2 storage processors

• Unisphere Manager 1.1 or later

• Off Array UI 1.1 or later • Host Agent 6.26 or later, including Unisphere 1.1 • Server Utility 1.1 or later • Initialization Utility 1.1 or later • Unisphere Service Manager 1.1 or

The following matrices can be found on http://Support.EMC.com

• VNX5100 and VNX5300 series Disk and OE Matrix (PN 300-012-417)

• VNX5500, VNX5700, and VNX7500 series Disk and OE Matrix (PN 300-012-419)

http://support.emc.com/

5

Fixed problems


Storage system type

Minimum hardware requirements

Minimum software requirements

Supported Disk/Flash Drives

later • CLI 1.1 or later

Fixed problems

This section contains fixed problems for 05.32.000.5.015.

Symptom description

Fix summary Solution

Platforms: VNX5100 VNX5300 VNX5500 VNX5700 VNX7500 Severity: Medium Frequency of occurrence: Likely under a rare set of circumstances Tracking number: 49796102/505950 47721910/487140 46752998/478875 46841320/478324

A single storage processor bugcheck (0x01901007) occurred when an attached host violated Fibre Channel protocol by sending overlapped commands.

Corrected code to better track overlapped commands to prevent this bugcheck. KnowledgeBase ID: None

Exists in versions: 05.32.000.5.006 05.32.000.5.008 05.32.000.5.011 Fixed in versions: 05.32.000.5.015

6

Fixed problems


Platforms: VNX5100 VNX5300 VNX5500 VNX5700 VNX7500 Severity: Medium Frequency of occurrence: Rarely under a rare set of circumstances Tracking number: 49535074/504303

A single storage processor reboot (0x00000041) occurred on a storage system with iSCSI connected layered applications.

Fixed the memory leak that led to the bugcheck. KnowledgeBase ID: None


Platforms: VNX5100 VNX5300 VNX5500 VNX5700 VNX7500 Severity: Medium Frequency of occurrence: Infrequently under a specific set of circumstances Tracking number: 503823

A single storage processor bugcheck (0x05900000) occurred when multiple trespasses were being processed for a RAID group, and in between the trespasses either a hot spare swapped in or a drive was replaced.

Changed the hot spare and new drive processing code to better handle queued trespasses. KnowledgeBase ID: None


Platforms: VNX5100 VNX5300 VNX5500 VNX5700 VNX7500

Pool LUNs went offline when a LUN was queued for trespass behind other LUNs in the same RAID group, and a backend failure (drive or enclosure

Changed the trespass code to properly handle this scenario. KnowledgeBase ID: None

Exists in versions: 05.32.000.5.006 05.32.000.5.008 05.32.000.5.011 Fixed in versions:

7

Fixed problems


Severity: Critical Frequency of occurrence: Infrequently under a specific set of circumstances Tracking number: 488458

failure) occurred.

05.32.000.5.015

Platforms VNX5100 VNX5300 VNX5500 VNX5700 VNX7500 Severity: Critical Frequency of occurrence: Rarely under a rare set of circumstances Tracking number: 510018

A transient hardware issue on a storage processor resulted in unexpected behavior.

Improved firmware detection and handling of transient hardware issue. KnowledgeBase ID: None



Symptom description


8

Fixed problems


Symptom description


Platforms VNX5100 VNX5300 VNX5500 VNX5700 VNX7500 Severity: Medium Frequency of occurrence: Always during a specific event Tracking number: 500833

In manual IPv6 mode, IPv6 confirmation information was cleared when IPv4 information was changed. This resulted in loss of IPv6 connectivity.

Corrected code to only modify requested data, leaving all other settings intact. KnowledgeBase ID: None

Exists in versions: 05.32.000.5.006 05.32.000.5.008 Fixed in versions: 05.32.000.5.011

Platforms VNX5100 VNX5300 VNX5500 VNX5700 VNX7500 Severity: Medium Frequency of occurrence: Always under a specific set of circumstances Tracking number: 48260182/492447

Unbind LUN failed because a drive was stuck in power up state.

Corrected code to handle the case when a drive fails during zero mark processing. KnowledgeBase ID: None


Platforms VNX5100 VNX5300 VNX5500 VNX5700 VNX7500

A storage processor bugcheck (0x000000d1) occurred when a drive returned a check condition.

Corrected code to properly handle the drive check condition. KnowledgeBase ID: None


9

Fixed problems


Symptom description


Severity: Medium Frequency of occurrence: Rarely under a rare set of circumstances Tracking number: 48464980/496715

Platforms: VNX5100 VNX5300 VNX5500 VNX5700 VNX7500 Severity: Medium Frequency of occurrence: Rarely under a specific set of circumstances Tracking number: 45943320/470897

During heavy I/O targeted to a single storage processor, the peer storage processor bugchecked (0x00000041).

Balanced internal algorithms to periodically give up control to allow other processes to run in this scenario. KnowledgeBase ID: None


10

Fixed problems


Symptom description


Platforms: VNX5100 VNX5300 VNX5500 VNX5700 VNX7500 Severity: Critical Frequency of occurrence: Likely under a rare set of circumstances Tracking number: 46985652/479789

A read failure during pool deletion caused the pool deletion to fail.

If a media error was encountered while processing a read request for a LUN in a RAID-1 or RAID-1/0 RAID group, and the first drive in the RAID group failed during the recovery from the media error, the read could fail. Corrected media error recovery code to handle this scenario and complete the read. KnowledgeBase ID: None


Platforms: VNX5100 VNX5300 VNX5500 VNX5700 VNX7500 Severity: Critical Frequency of occurrence: Likely under a specific set of circumstances Tracking number: 47259290/482353

Host I/O timed out due to drive errors.

Enhanced code that avoids full rebuilds during single path failures to allow rebuilds to begin immediately for drive failures that indicate the failure will repeat on the peer side. KnowledgeBase ID: None


Platforms: VNX5100 VNX5300 VNX5500 VNX5700 VNX7500

LUN compression hung when the destination LUN/pool was faulted.

Corrected internal code logic. KnowledgeBase ID: None


11

Fixed problems


Symptom description


Severity: Medium Frequency of occurrence: Likely under a specific set of circumstances Tracking number: 494493

Platforms: VNX5100 VNX5300 VNX5500 VNX5700 VNX7500 Severity: Medium Frequency of occurrence: Likely under a specific set of circumstances Tracking number: 494492

Storage processor bugcheck (0x05900000) occurrred when starting LUN migration.

Corrected internal code logic KnowledgeBase ID: None


12

Fixed problems


Symptom description


Platforms: VNX5100 VNX5300 VNX5500 VNX5700 VNX7500 Severity: Medium Frequency of occurrence: Likely under a specific set of circumstances Tracking number: 49290924/501471

After a 05.31 to 05.32 NDU failed, the automatic revert to 05.31 failed, leaving one storage processor out of service.

Changed software packaging to avoid leaving incompatible drivers when a 05.31 to 05.32 NDU fails at a particular point in the upgrade. KnowledgeBase ID: None


Platforms: VNX5100 VNX5300 VNX5500 VNX5700 VNX7500 Severity: Critical Frequency of occurrence: Rarely under a specific set of circumstances. Tracking number: 493415

Non-disruptive upgrade failed with bugcheck (various, including 0x71518026). It reverted successfully back to the original software version.

Fixed the stack overflow issue that led to the bugcheck. KnowledgeBase ID: None


13

Fixed problems



Brief description Symptom details Solution Platforms: VNX5100 VNX5300 VNX5500 VNX5700 VNX7500 Severity: Critical Frequency of occurrence: Likely under a specific set of circumstances Tracking number: 48780368/499021 46946692/481912 48151092/491189

Storage processor rebooted unexpectedly (various codes, including 0xe117b264, 0x0000007e and 0x00000044) if Recoverpoint splitter was being used

Recoverpoint splitter did not correctly handle IO cancellation requests, resulting in the storage processor rebooting. KnowledgeBase ID: None

Exists in versions: 05.31.000.5.006 Fixed in version: 05.32.000.5.008

Platforms: VNX5100 VNX5300 VNX5500 VNX5700 VNX7500 Severity: Critical Frequency of occurrence Likely under a specific set of circumstances Tracking number: 500729

Storage processor rebooted unexpectedly when a Recoverpoint consistency group was disabled

Under heavy IO load the splitter could cause the storage processor to reboot when a Recoverpoint consistency group was disabled. KnowledgeBase ID: None

Exists in versions: 05.31.000.5.006 Fixed in version: 05.32.000.5.008

14

Fixed problems



Brief description Symptom details

Solution

Platforms: VNX5100 VNX5300 VNX5500 VNX5700 VNX7500 Severity: Medium Frequency of occurrence: Likely under a specific set of circumstances. Tracking number: 36361816/380605

DAE fan faults were reported against the wrong fan.

DAE fan faults were reported against the wrong fan. For example, fan A in enclosure 0-0 was reported as faulted when the failed fan was fan B in enclosure 2-5. KnowledgeBase ID: None

Exists in versions: 05.31.000.5.006 05.31.000.5.007 05.31.000.5.508 05.31.000.5.011 05.31.000.5.012 05.31.000.5.704 05.31.000.5.709 05.31.000.5.716 05.31.000.5.720 Fixed in versions: 05.32.000.5.006

Platforms: VNX5100 VNX5300 VNX5500 Severity: Low Frequency of occurrence: Always under a specific set of circumstances. Tracking number: 399655

If the maximum number of supported Fibre Channel ports for a VNX array was exceeded, removing the module did not clear the over-limit fault.

If a Fibre Channel module was inserted that exceeded the maximum number of supported Fibre Channel ports for a VNX array, removing the module did not clear the over-limit fault. KnowledgeBase ID: None

Exists in versions: 05.31.000.5.006 05.31.000.5.007 05.31.000.5.008 05.31.000.5.011 05.31.000.5.012 05.31.000.5.502 05.31.000.5.509 05.31.000.5.704 05.31.000.5.709 05.31.000.5.716 05.31.000.5.720 Fixed in versions: 05.32.000.5.006

Platforms: VNX5100 VNX5300 VNX5500 VNX5700 VNX7500 Severity: Medium

A single storage processor bugcheck occured on the secondary array in a MirrorView™

connection.

A single storage processor bugcheck occured on the secondary array in a MirrorView connection during recovery from a failed non-disruptive upgrade. This occurred

Exists in versions: 05.31.000.5.006 05.31.000.5.007 05.31.000.5.008 05.31.000.5.011 05.31.000.5.012 05.31.000.5.502 05.31.000.5.509 05.31.000.5.704

15

Fixed problems



Solution

Frequency of occurrence: Likely under a specific set of circumstances. Tracking number: 388709

because the mirrors were not fractured before the software upgrade. KnowledgeBase ID: None

05.31.000.5.709 05.31.000.5.716 05.31.000.5.720 Fixed in versions: 05.32.000.5.006

Platforms: VNX5300 VNX5500 VNX5700 VNX7500 Severity: Medium Frequency of occurrence: Likely under a specific set of circumstances. Tracking number: 401751

A single storage processor bugcheck occured when the flow control setting on an iSCSI port was repeatedly changed.

A single storage processor bugcheck occured when the flow control setting on an iSCSI port was repeatedly changed. KnowledgeBase ID: None

Exists in versions: 05.31.000.5.006 05.31.000.5.007 05.31.000.5.008 05.31.000.5.011 05.31.000.5.012 05.31.000.5.502 05.31.000.5.509 05.31.000.5.704 05.31.000.5.709 05.31.000.5.716 05.31.000.5.720 Fixed in versions: 05.32.000.5.006

Platforms: VNX5100 VNX5300 VNX5500 VNX5700 VNX7500 Severity: Medium Frequency of occurrence: Rare under a specific set of circumstances. Tracking number: 408996

LED fault light was lit, but the Unisphere status did not properly report which LCC was at fault.

When an LCC in a 60-drive DAE experienced certain fault conditions, the LED fault light was lit, but the Unisphere status did not properly report which LCC was at fault. KnowledgeBase ID: None

Exists in versions: 05.31.000.5.502 05.31.000.5.509 05.31.000.5.704 05.31.000.5.709 05.31.000.5.716 05.31.000.5.720 Fixed in versions: 05.32.000.5.006

16

Known problems and limitations



Solution

Platforms: VNX5100 VNX5300 VNX5500 VNX5700 VNX7500 Severity: Medium Frequency of occurrence: Always under a specific set of circumstances. Tracking number: 431773

If an LCC in a DAE 60 failed, Unisphere status reported the drives connected to that LCC as removed and faulted.

If an LCC in a DAE 60 failed, Unisphere status reported the drives connected to that LCC as removed and faulted even though they were operating properly through the alternate port. KnowledgeBase ID: None

Exists in versions: 05.31.000.5.502 05.31.000.5.509 05.31.000.5.704 05.31.000.5.709 05.31.000.5.716 05.31.000.5.720 Fixed in versions: 05.32.000.5.006


This section includes all of the known problems found in 05.32.000.5.015.

Brief description Symptom details Solution (or workaround)

Platforms: VNX5300 VNX5500 VNX5700 VNX7500 Severity: Medium Frequency of occurrence: Rarely under a specific set of circumstances Tracking number: 383572

10 Gb iSCSI SLICs might lose connectivity.

10 Gb iSCSI SLICs might lose connectivity because of a hardware failure on the module, leading to hosts logging out, but there is no event log message indicating a hardware failure. Note: This symptom might be seen for other reasons and requires analysis to determine the true cause.

Contact your service provider to determine if the hardware should be replaced. Exists in versions: 05.32.000.5.006 05.32.000.5.008 05.32.000.5.011 05.32.000.5.015

Platforms: VNX5100 VNX5300

Logical units might not enable following drive shutdown in a

To protect against over-temperature, if one DAE power supply is physically

Do not remove the power supply for more than 5 minutes. If this error is

17




VNX5500 VNX5700 VNX7500 Severity: Critical Frequency of occurrence: Likely under a specific set of circumstances. Tracking Number: 395579

DAE.

removed for longer than 5 minutes, all the drives in the DAE are shut down. When the power supply is replaced, logical units bound on pools using drives in that enclosure might not enable.

encountered, remove and reconnect the cable for this enclosure. If this does not resolve the problem, contact your service provider. Exists in versions: 05.32.000.5.006 05.32.000.5.008 05.32.000.5.011 05.32.000.5.015

Platforms: VNX5300 VNX5500 VNX5700 VNX7500 Severity: Medium Frequency of occurrence: Likely under a specific set of circumstances. Tracking number: 397450

If both storage processors are rebooted at the same time, one storage processor might report SFP faults on the FCoE module.

On a VNX array running FCoE, if both storage processors are rebooted at the same time, one storage processor might report SFP faults on the FCoE module. The fault occurs because of contention between the two storage processors.

Re-seat the faulted SFPs, or reboot the storage processor with the faults. Exists in versions: 05.32.000.5.006 05.32.000.5.008 05.32.000.5.011 05.32.000.5.015

18




Platforms: VNX5100 VNX5300 VNX5500 VNX5700 VNX7500 Severity: Low Frequency of occurrence: Rarely Tracking number: 463796

Single storage processor reboot during normal operation.

Bugcheck 0x0000000002503021

System recovers automatically. Exists in versions: 05.32.000.5.006 05.32.000.5.008 05.32.000.5.011 05.32.000.5.015

Platforms: VNX5100 VNX5300 VNX5500 VNX5700 VNX7500 Severity: Low Frequency of occurrence: Rarely under a specific set of circumstances. Tracking number: 485084

Unable to persist a new I/O module.

Error message when sending I/O to new SLIC: “'Port not yet available for configuration”.

Reboot the SP. Exists in versions: 05.32.000.5.006 05.32.000.5.008 05.32.000.5.011 05.32.000.5.015

The following section lists issues that might affect the configuration or repair of the storage system.

Table 3 lists the maximum number of LUNs and drives/enclosures supported per platform.

19



Table 3: LUNs and drive/enclosures supported Platform Maximum

number of LUNs Maximum number of drives/enclosures

VNX5100 512 75 VNX5300 2048 125 VNX5500 4096 250 VNX5700 4096 500 VNX7500 8192 1000

All VNX systems support a maximum of ten enclosures per back-end port. Enclosures added that exceed this limit, or that exceed the maximum drive slot count for that system, are not allowed to power on by the storage system.

Non-disruptive software upgrade (NDU)

Do not perform a software upgrade (NDU) when the total CPU utilization of both storage processors is greater than 100 percent. Factors that may impact this include:

Heavy I/O activity

Large numbers of LUNs on VNX for Block database drives (slots 0_0, 0_1, 0_2, or 0_3)

Heavy administrative activity

If your storage system exceeds this limit, the Acceptable Processor Utilization pre-software upgrade rule will prevent an upgrade from proceeding. This will not cause any disruption to the storage system, but it might cause a scheduled software upgrade to be delayed until the storage system is less busy.

Replication software configurability

Remote replication over the FCoE ports is not supported. Remote replication works on a storage system with FCoE, but the replication works only over the iSCSI and Fibre Channel ports.

Restrictions during non-disruptive software upgrade

You cannot initiate any bind or expansion operations while an NDU is in progress. The software will reject any attempts to do so.

Software commits

After performing a software upgrade (NDU), you might not be able to

20



bind or unbind LUNs or create or destroy RAID groups or storage pools until you have committed the new software. Once the software is committed, you cannot revert to the previous revision of the software. You can revert only before the software is committed. Additionally, you must commit the currently active software before attempting an NDU of a new package.

When performing an NDU commit, do not execute any configuration administration commands until the commit is complete. This includes binding /unbinding LUNs, creating/destroying RAID groups, creating/destroying mirrors, adding/removing secondary mirrors, allocating/de-allocating the remote mirror write intent log, creating/deleting a snapshot, adding/removing a LUN to/from the snap cache, adding/deleting SnapView™ clones, or allocating/de-allocating snap clones private LUNs.

LUN migration

The following limitations apply to LUNs:

EMC does not support the migration of a component LUN that is part of an aggregate metaLUN.

The destination LUN must be a public LUN when it is chosen for migration (this is the normal case).

No other layered feature can use a destination LUN.

The destination LUN cannot be in a user-defined storage group.

Migration and LUN expansion are mutually exclusive operations. You cannot migrate an expanding LUN, and you cannot expand a migrating LUN.

The source LUN cannot be a private LUN. If a layered application uses the source LUN, you can migrate the

source LUN only to a LUN of the same size.

After a migration completes, the data reported in the inquiry command will still reflect the LUN's original RAID type. The new RAID type is visible through the Unisphere interface. This behavior is as designed and is required to maintain compatibility with some operating systems.

If the destination LUN is a thin LUN, it may not be fully provisioned after a migration because the thin LUN is enabled for zero detection.

The maximum number of migrations that can be executing at one time is as follows:

VNX5100 8 VNX5300, VNX5500 16 VNX5700, VNX7500 24

21

Technical notes


RAID group expansion

VNX for Block software does not support RAID group expansion.

LUN shrink

VNX for Block software does not support shrinking the size of LUNs, except for Virtual Provisioning™ LUNs.

10 Gb host connectivity

On a Windows host with one or more Broadcom 57710 10 gigabit network interface cards (NICs) and one or more NICs from another vendor, if the Broadcom Advanced Server Program is used to add VLAN tags to a Broadcom NIC, those VLAN tags are also added to the non-Broadcom NICs. This might result in lost network connectivity for the non-Broadcom NICs, since the network infrastructure to which it is connected might not be configured to use that VLAN tag.

Technical notes

This section includes technical information for the VNX Operating Environment for Block software.

NDU

Do not perform a software upgrade on a system that has hardware faults.

Storage processor startup

EMC recommends that you reboot a storage processor only during periods of light I/O activity.

Disk firmware upgrades

The online disk firmware upgrade (ODFU) wizard checks for disk firmware update recommendations automatically. This allows you to download the latest disk firmware from http://Support.EMC.com and install it without disruption to storage system operation.

Soft logical sector data errors

A soft logical sector data error indicates that the data redundancy information for a sector stored on a redundant LUN is not valid. These errors occur because a power failure or disk failure interrupted a write in


22

Technical notes


progress. VNX for Block software fully recovers from these errors, and reports them to the event log only for informational purposes.

When the storage system returns to a fully functional state, the sniff verify utility automatically finds all redundant data inconsistencies and takes the necessary steps to correct them. A write to a sector with inconsistent redundant information also corrects the inconsistency.

Note: These errors are normal in the event of a power failure.

Event log entries appear out of order

VNX OE for Block software may not always enter items into the event log in the order that they occurred. This happens only when several events occur at approximately the same time. Time stamps are correct.

Hot spare ownership

Once a hot spare is assigned to a RAID Group, that RAID Group owns it until the RAID Group is repaired. If the RAID Group is accidentally double-faulted because you removed the wrong disk during repair, then re-inserting the incorrectly removed disk will cause the RAID Group to return to its original state (including continued operation with the hot spare engaged).

Disk insertion and removal

When inserting or removing more than one disk from a storage system, you must pause between each insertion and removal to allow the storage system to recover. Before inserting or removing the next disk module, make sure that:

A newly inserted disk's amber light is off and its green light has returned to a normal state: on or blinking due to I/O in progress.

On a disk removal, the green lights for the other disks return to a normal state: on or blinking due to I/O in progress.

Database disk replacement

Replace a disk in slots 0_0, 0_1, 0_2, or 0_3 of DPE enclosure 0 only when the storage system is fully operational. The disks in these slots are VNX OE for Block software database disks. Each database disk contains an image of the storage-system configuration and current operating status. VNX OE for Block software ensures that the newly inserted disk has the correct database information copied to it when it is installed while the storage system is running. If you insert a disk into one of these slots when the storage system is powered off, the correct VNX OE for Block

23

Technical notes


software database information is not written to the replaced database disk.

This can result in storage system configuration problems when the storage system is powered up or if another database disk fails. If you install the VNX OE for Block software database disk and you find that existing LUNs are not enabled or caching cannot be enabled or disabled, contact your service provider.

Software image disk replacement

The VNX OE for Block software image is stored on disks 0_0, 0_1, 0_2, and 0_3. Disks 0_0 and 0_2 form a mirror with storage processor A's software, and disks 0_1 and 0_3 form a mirror with storage processor B's software. You should never remove and/or replace both disks in one of the mirrors at the same time. If you do, no software will exist to support that storage processor, and it will become nonfunctional.

Extended inquiry information

Extended inquiry information is not available when running with DMP failover mode in a fibre open initiator type. This means that no vendor-specific information will be returned on an inquiry request in this mode.

Faulted unbound drive

If a disk drive is in a faulted state but not part of a bound LUN, then the drive fault LED will be lit and a log entry will indicate the drive is faulted. Unisphere will not display the drive as faulted, but will show the enclosure as faulted. If the drive is part of a bound LUN, then Unisphere will show the drive as faulted.

LCC failure management

VNX OE for Block software is designed to bugcheck with a 0x00800094 code if a storage processor cannot access any disks on its back-end loop. This initiates a trespass of this storage processor's LUNs to the peer storage processor, ensuring the integrity of user data. This bugcheck is a very unlikely event and typically only an LCC hardware failure causes it.

Following the bugcheck, the storage processor attempts to reboot several times, and fails to completely initialize as long as this hardware failure is present. In this situation, the storage processor remains in a shutdown state.

After the LCC is repaired, you must power-cycle the storage system or remove the storage processor that bugchecked and reinsert it (effectively

24

Technical notes


power-cycling the storage processor) to allow the storage processor to reboot and return to an operational state.

Storage processor does not power up

If both SPE fans are unplugged when you power up a storage processor, the storage processor does not come up until you plug in the fans.

IP addressing

If the storage system is configured with an invalid IP address/subnet mask/gateway combination (for example, the gateway is not on the same subnet as the storage system), then the gateway value will not be set, nor will it be displayed.

Ensure that valid combinations are entered for the IP parameters. For example, the following parameters are invalid because the gateway (10.5.1.1) is not reachable from the local subnet (10.5.2) and will not be set.

IP address: 10.5.2.155

Subnet mask: 255.255.255.0

Gateway: 10.5.1.1

A proper value for the gateway for the IP address entered would be 10.5.2.1.

EMC UltraFlex I/O modules

I/O modules and SFPs can be added and replaced. You cannot upgrade I/O modules to a different type. You can add I/O modules, but a service provider must replace I/O modules.

You cannot delete ports.

Default rebuild priority

The rebuild priority default value when binding a LUN is High for this version of software.

Default verify priority

The verify priority default value when binding a LUN is Medium for this version of software.

25

Documentation


I/O module replacements

This version of VNX OE for Block does not support replacing a 10 Gbps with a 1 Gbps I/O module. If you do this, the SLIC will not power up and the storage processor where the I/O module was replaced might become unmanageable.

Switch requirements for 8 Gbps Fibre Channel I/O module

If the 8 Gbps I/O module is connected to an 8 Gbps Connectrix B series switch (Models DS-300B, DS-5100B, or DS-5300B), the switch will require firmware version FOS 6.1.0j or later so that auto-negotiation will work correctly. As a workaround, you can manually set the switch ports to 8 Gbps.

Documentation

Table 4 lists the documents applicable for this version of the product.

Table 4: Available documentation Part Number Description

300-013-472 EMC Unisphere Release Notes 300-013-473 EMC Unisphere Host Agents/CLI and Utilities Release Notes

Software media, organization, and files

This section does not apply to VNX Operating Environment for Block software.

Installation

The VNX Operating Environment for Block is pre-installed in the factory or is available in the upgrade bundle. Optional software components are enabled by purchasing and installing the appropriate enabler package.

To install enablers or upgrade VNX Operating Environment for Block, use the wizards in the System Software section of the Unisphere Service Manager (USM). The USM replaces the Navisphere Service Taskbar (NST) and is available from the VNX Software Update Tools page at http://Support.EMC.com/.

Refer to the information on software commits in the Known problems and limitations section in this document for additional information


26

Troubleshooting and getting help


relevant to software upgrades.

Installation restriction

The following restriction applies to the VNX5100, VNX5300, VNX5500, VNX5700, and VNX7500 platforms:

When installing a new version of VNX Operating Environment for Block software, you must not remove a storage processor or reboot the storage system while the software installation is in progress. Doing so might leave the storage system in a transient state from which it cannot be restarted. Wait for the entire installation process to complete and automatically reboot.

Installation procedure

EMC recommends limiting software installation (NDU) to periods of light input and output activity.


This section provides specific instructions for contacting EMC Customer Support and obtaining additional information about EMC products.

Where to get help

EMC support, product, and licensing information can be obtained as follows:

Product information For documentation, release notes, software updates, or for information about EMC products, licensing, and service, go to EMC Online Support (registration required) at: http://Support.EMC.com.

Troubleshooting Go to EMC Online Support. After logging in, locate the appropriate Support by Product page.

Technical support For technical support and service requests, go to EMC Customer Service on EMC Online Support. After logging in, locate the appropriate Support by Product page and choose either Live Chat or Create a service request. To open a service request through EMC Online Support, you must have a valid support agreement. Contact your EMC Sales Representative for details about obtaining a valid support agreement or to answer any questions about your account.


27



Note: Do not request a specific support representative unless one has been assigned to your particular system problem.

28



Copyright © 2012 EMC Corporation. All Rights Reserved.

Published December 2012

EMC believes the information in this publication is accurate as of its publication date. The information is subject to change without notice.

THE INFORMATION IN THIS PUBLICATION IS PROVIDED "AS IS." EMC CORPORATION MAKES NO REPRESENTATIONS OR WARRANTIES OF ANY KIND WITH RESPECT TO THE INFORMATION IN THIS PUBLICATION, AND SPECIFICALLY DISCLAIMS IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.

Use, copying, and distribution of any EMC software described in this publication requires an applicable software license.

EMC2, EMC, and the EMC logo are registered trademarks or trademarks of EMC Corporation in the United States and other countries. All other trademarks used herein are the property of their respective owners.

For the most up-to-date regulatory document for your product line, go to the technical documentation and advisories section on the EMC online support website.

EMC VNX FAST VP for VNX OE for Block Release Notes · PDF file03/12/2012 · 1 EMC® VNX Operating ... VNX Block to Unified upgrade This version of VNX OE for Block supports ... EMC

Documents