Pillar Axiom 500 Storage System - Version All Versions to All Versions [Release All Releases]
Pillar Axiom 600 Storage System - Version All Versions to All Versions [Release All Releases]
This document provides the necessary steps to generate the back-end topology of a Pillar Axiom 500 and Axiom 600.
The tool Topology.exe attached to that document has been tested on the following platforms with AxiomONE R4 and R5:
Windows 7 64bit
Windows Server 2008 R2 64bit
Windows XP SP3 32bit
Windows Server 2003 32bit
The tool checks for the following items:
- SFP in BYPASSED mode.
- SFP from manufacturers different to the ones used by manufacturing.
- Invalid connections.
- Mix of optical and copper connections.
- Ports using the wrong speed.
- Ports on the Private Interconnect Module that should not be used to connect bricks.
- Inconsistencies between string paths.
- Missing crossover brick connections.
- Invalid and missing inter-slammer / intra-slammer connections.
The tool can be useful when installing or troubleshooting a new Axiom as it provides the topology in less than a minute. It can also provide additional details when the GUI reports an Invalid Topology.
Open the command prompt or double click on Topology.exe and provide the management IP address or DNS name of the Axiom:
|
D:\>Topology.exe -h
Topology.exe (designed for Pillar AxiomONE R4 & R5)
Description: Generate the Axiom back-end topology into a text file.
Usage: Topology.exe -i <Axiom IP Address or DNS name> [-u <user name>] [-p <pass
word>] [-o <output.txt>]
-i | --AxiomIP <Axiom IP Address or DNS name to query>
[-u | --UserName <user name account on the Axiom>]
[-p | --password <password>]
[-o | --outputfile <file to be generated by Topology.exe>]
D:\>Topology.exe -i 10.10.10.10
The output file has been successfully created: A002497BLU_Topology_2013-12-19T19_04_33.txt
D:\>
|
Open the generated text file with notepad or any text editor.
Known limitation:
You might see this message if you have a configuration with SATA brick first generation:
There is a mix of SATA legacy bricks and other brick types on that string. This is not recommended for performance and troubleshooting purposes.
FCR, SATA V2 and FCR V2 physically route Fibre-Channel signals differently than the SATA v1 Bricks.
Having two different signals on the same loop can make back-end troubleshooting very difficult in terms of which component(s) is generating the errors.
It also creates an issue to determine the exact order of the Bricks in the string: as we have two different signals, we also get two groups of Bricks.
The first group has SATA brick(s) and the second has at least one or more bricks of these brick types: FCR/SATA V2/FCR V2.
The order of each group is provided by Topology.exe but it is not possible to know the order of bricks within the two groups.
Example:
(a) FS1 <-- Brick001 (*583A - SATA) - Brick002 (*5883 - SATA) - Brick00A (*672A - FCR V2) - Brick00B (*672D - FCR V2)
The following statements are true:
- Brick001 is upstream compared to Brick002
- Brick00A is upstream compared to Brick00B
The following statements are unknown:
- Brick001 is upstream compared to Brick00A and Brick00B
- Brick00A is upstream compared to Brick001 and Brick002
Here are the 5 other potential scenarios:
(b) FS1 <-- Brick001 (*583A - SATA) - Brick00A (*672A - FCR V2) - Brick002 (*5883 - SATA) - Brick00B (*672D - FCR V2)
(c) FS1 <-- Brick001 (*583A - SATA) - Brick00A (*672A - FCR V2) - Brick00B (*672D - FCR V2) - Brick002 (*5883 - SATA)
(d) FS1 <-- Brick00A (*672A - FCR V2) - Brick00B (*672D - FCR V2) - Brick001 (*583A - SATA) - Brick002 (*5883 - SATA)
(e) FS1 <-- Brick00A (*672A - FCR V2) - Brick001 (*583A - SATA) - Brick00B (*672D - FCR V2) - Brick002 (*5883 - SATA)
(f) FS1 <-- Brick00A (*672A - FCR V2) - Brick001 (*583A - SATA) - Brick002 (*5883 - SATA) - Brick00B (*672D - FCR V2)
There would more or less scenarios depending on the number of bricks in the string.
The algorithm used by Topology.exe checks the Brick type of the last member of the FCR/SATA V2/FCR V2 group and see if the Brick has downstream bricks (it would be SATA v1) -> it checks the Brick RC0 FC1 and RC1 FC1 ports connectivity.
If there is nothing connected, the SATA bricks are listed first and the FCR/SATA V2/FCR V2 are printed at the end of the string. Otherwise, we print the SATA at the end of the string.
Always perform a physical inspection of the cabling. You should start the inventory from the Slammer FS port of the string with a mix of components and Identify Slammers and Bricks using the GUI.
One exception where the mix can be reliable:
FS2 <-- Brick006 (*7DAC - SATA) - Brick014 (*7A7D - SATA) - Brick019 (*7BB2 - SATA V2) <-- there is only one "FCR/SATA V2/FCR V2" brick and we know that the FC1 ports are not connected.
However, it only works if the last brick is a SATA V2 or FCR V2 as FCR bricks always report the downstream port as Connected.
On the above (a), (b) and (f) examples, we can say that Brick00B is indeed at the end of the String but we do not know the real position of Brick00A.
One of the Cabling Best Practices is to wire SATA v1 bricks on separate strings.
Reference: Axiom cabling guide.
[This section is not visible to customers.]
The script Topology.pl is available on the log servers. It has been tested on R3, R4 and R5.
Run the script on the main folder.
|
-bash-4.1$ Topology.pl -h
Topology.pl (designed for Pillar AxiomONE R4 & R5)
Description: Generate the Axiom back-end topology into a text file.
The tool needs to be run on the main log folder.
Usage: Topology.pl [-o <Topology_A009999ZZZ.txt>]
[-o | --outputfile <file to be generated by Topology.pl>]
-bash-4.1$ Topology.pl
The output file has been successfully created: Topology_A002816BLC.txt
|
And open the generated file with any file viewer application.
Regarding “We cannot confirm that the FC3 <-> FS10 connections are within the same node”:
The collected logs on R5 do not have information regarding that specific connection (it only applies on PIM v2). You should advise the customer to run Topology.exe to get the real status of that connection if they recently had a new Slammer installation or recabled their Axiom. We have seen people making these types of mistake: using FS11 instead of FS10, connecting to FS10 on the opposite CU instead of the same CU, or simply forgetting to connect FC3 <-> FS10 because they were looking at the wrong paragraph of the cabling guide).
Examples of detected issues:
SFP/cable issue:
|
***********
* WARNING *
***********
BRICK-010 RC0 FC0 SFP status is BYPASSED_LOSS_OF_SYNC.
BRICK-010 RC0 FC3 SFP status is BYPASSED_LOSS_OF_SYNC.
BRICK-002 RC1 FC1 SFP status is BYPASSED_LOSS_OF_SYNC.
|
Mix of copper and optical cables:
|
***********
* WARNING *
***********
Please engage Oracle Support to review the cabling of the BackEnd (mix of Copper & Optical connectors).
BRICK-019 RC0 FC0 SFP is from AVAGO.
BRICK-019 RC0 FC2 SFP is from Molex Inc.
BRICK-019 RC0 FC3 SFP is from AVAGO.
BRICK-019 RC1 FC0 SFP is from AVAGO.
BRICK-019 RC1 FC2 SFP is from Molex Inc.
BRICK-019 RC1 FC3 SFP is from AVAGO.
BRICK-020 RC0 FC0 SFP is from AVAGO.
BRICK-020 RC0 FC2 SFP is from Molex Inc.
BRICK-020 RC0 FC3 SFP is from AVAGO.
BRICK-020 RC1 FC0 SFP is from AVAGO.
BRICK-020 RC1 FC2 SFP is from Molex Inc.
BRICK-020 RC1 FC3 SFP is from AVAGO.
BRICK-019 RC0 FC3 speed is 4Gb/s instead of 2Gb/s.
BRICK-019 RC1 FC3 speed is 4Gb/s instead of 2Gb/s.
BRICK-020 RC0 FC3 speed is 4Gb/s instead of 2Gb/s.
BRICK-020 RC1 FC3 speed is 4Gb/s instead of 2Gb/s.
|
SFPs not provided by manufacturing (we support Molex & Amphenol for copper and AVAGO, FINISAR & CISCO-FINISAR for optical):
|
***********
* WARNING *
***********
Brick009 RC0 FC2 SFP is from BROCADE.
Brick009 RC1 FC2 SFP is from BROCADE.
Brick020 RC1 FC2 SFP is from AGILENT.
Slammer1 CU0 FS4 SFP is from BROCADE.
Slammer1 CU1 FS1 SFP is from BROCADE.
|
Single Slammer configuration where ports are used to connect bricks (not supported):
|
***********
* WARNING *
***********
Slammer1 CU0 FS7 should not be connected (not supported).
Slammer1 CU1 FS5 should not be connected (not supported).
...
Slammer1 CU0 (*5622)
...
FS7 <-- Brick001 (*66FB - SATA)
Slammer1 CU1 (*562A)
FS2 <-- Brick002 (*66FD - SATA)
FS3 <-- Brick003 (*6C91 - SATA)
FS4 <-- Brick001 (*66FB - SATA)
FS5 <-- Brick004 (*6C7E - SATA) - Brick006 (*6C7A - SATA) - Brick005 (*6C78 - SATA)
|
Mix of SATA v1 and other brick types:
|
SLAMMER-02 CU1 (*378A)
FS4 <-- BRICK-002 (*5F6D - SATA) - BRICK-011 (*755F – SATA V2) - BRICK-019 (*7A59 - SATA V2) (1)
***********
* NOTE *
***********
(1) There is a mix of SATA legacy bricks and other brick types on that string. This is not recommended for
performance and troubleshooting purposes.
In that case, we cannot guarantee that the Bricks are in order, please see KM 1609365.1 for more details.
|
Brick installed with RC0 connected to Slammer CU1 and RC1 connected to Slammer CU0:
|
Slammer1 CU0 (*7C12)
FS4 <-- Brick001 (*7690 - SATA V2) (1)
Slammer1 CU1 (*7C1A)
FS2 <-- Brick001 (*7690 - SATA V2) (2)
***********
* WARNING *
***********
(1) This string is connected to the wrong Slammer Control Unit: it should be CU1.
(2) This string is connected to the wrong Slammer Control Unit: it should be CU0.
|
Missing a connection between two bricks on the first path:
|
SLAMMER-01 CU0 (*A4D2)
FS2 <-- BRICK-005 (*85AF - FCR V2) - BRICK-013 (*859C - FCR V2) (1)
...
SLAMMER-02 CU1 (*A9EA)
FS2 <-- BRICK-005 (*85AF - FCR V2) - BRICK-013 (*859C - FCR V2) - BRICK-021 (*7F6C - SATA V2)
***********
* WARNING *
***********
(1) This string is missing the connection between BRICK-013 RC0 FC1 and BRICK-021 RC0 FC2.
|
Cabling mistake:
|
Slammer1 CU0 (*5622)
FS1 <-- Brick003 (*6C91 - SATA)
FS2 <-- Brick002 (*66FD - SATA)
FS3 <-- Brick004 (*6C7E - SATA) - Brick005 (*6C78 - SATA) - Brick006 (*6C7A - SATA) (2)
FS4 <-- Brick001 (*66FB - SATA)
Slammer1 CU1 (*562A)
FS1 <-- Brick004 (*6C7E - SATA) - Brick006 (*6C7A - SATA) - Brick005 (*6C78 - SATA) (1)
FS2 <-- Brick001 (*66FB - SATA)
FS3 <-- Brick003 (*6C91 - SATA)
FS4 <-- Brick002 (*66FD - SATA)
***********
* WARNING *
***********
(1) Brick006 should not be connected to Brick004 RC1 FC1.
(2) Brick005 should not be connected to Brick004 RC0 FC1.
|
Missing a crossover connection:
|
SLAMMER-02 CU0 (*A7F2)
FS4 <-- BRICK-002 (*8493 - FCR V2) - BRICK-010 (*8486 - FCR V2) (1)
***********
* WARNING *
***********
(1) Missing connection from BRICK-010 RC0 FC0 to BRICK-010 RC1 FC3.
|
Missing a regular path:
|
SLAMMER-01 CU0 (*A232)
FS1 <-- BRICK-003 (*765E - SATA V2) - BRICK-011 (*7972 - SATA V2)
FS2 <-- BRICK-002 (*765D - SATA V2) - BRICK-010 (*764D - SATA V2) (1)
FS3 <-- BRICK-004 (*7661 - SATA V2) - BRICK-012 (*764B - SATA V2)
FS4 <-- BRICK-001 (*79A6 - SATA V2) - BRICK-009 (*79A5 - SATA V2)
...
SLAMMER-02 CU1 (*A14A)
FS1 <-- BRICK-004 (*7661 - SATA V2) - BRICK-012 (*764B - SATA V2)
FS2 <-- BRICK-001 (*79A6 - SATA V2) - BRICK-009 (*79A5 - SATA V2)
FS3 <-- BRICK-003 (*765E - SATA V2) - BRICK-011 (*7972 - SATA V2)
***********
* WARNING *
***********
(1) This string is missing the other path on SLAMMER-02 CU1 because BRICK-002 RC1 FC2 is not connected.
|
Another example of a missing path:
|
Slammer1 CU0 (*4812)
FS2 <-- BRICK-006 (*7FD0 - SATA V2) - Brick010 (*69B8 - SATA V2) - Brick014 (*69CE - SATA V2)
FS3 <-- BRICK-008 (*8009 - SATA V2) - Brick012 (*67F0 - SATA V2)
FS4 <-- BRICK-005 (*800F - SATA V2) - BRICK-009 (*7FC8 - SATA V2) - Brick013 (*69C4 - SATA V2)
Slammer1 CU1 (*481A)
FS1 <-- BRICK-008 (*8009 - SATA V2) - Brick012 (*67F0 - SATA V2)
FS2 <-- BRICK-005 (*800F - SATA V2) - BRICK-009 (*7FC8 - SATA V2) - Brick013 (*69C4 - SATA V2)
FS3 <-- Brick016 (*6B73 - FCR) / Brick015 (*6A4F - FCE) - Brick011 (*6812 - SATA V2) (2)
FS4 <-- BRICK-006 (*7FD0 - SATA V2) - Brick010 (*69B8 – SATA V2) - Brick014 (*69CE - SATA V2)
***********
* WARNING *
***********
(2) This string is missing the other path that is supposed to be on Slammer1 CU0.
|
Mistakes on the slammer interconnection cabling:
|
Slammer Interconnections:
SLAMMER-01 CU0 FC0 <-> SLAMMER-02 CU0 FS7 SLAMMER-01 CU1 FC0 <-> SLAMMER-02 CU1 FS7
SLAMMER-01 CU0 FC1 <-> SLAMMER-01 CU1 FS8 SLAMMER-01 CU1 FC1 <-> SLAMMER-01 CU0 FS8
SLAMMER-01 CU0 FC2 <-> SLAMMER-02 CU1 FS9 SLAMMER-01 CU1 FC2 <-> SLAMMER-02 CU0 FS10
SLAMMER-01 CU0 FC3 <-> SLAMMER-01 CU0 FS10 SLAMMER-01 CU1 FC3 <-> SLAMMER-01 CU1 FS10
SLAMMER-02 CU0 FC0 <-> SLAMMER-01 CU1 FS7 SLAMMER-02 CU1 FC0 <-> SLAMMER-01 CU0 FS7
SLAMMER-02 CU0 FC1 <-> SLAMMER-02 CU1 FS8 SLAMMER-02 CU1 FC1 <-> SLAMMER-02 CU0 FS9
SLAMMER-02 CU0 FC2 <-> SLAMMER-01 CU0 FS9 SLAMMER-02 CU1 FC2 <-> SLAMMER-01 CU1 FS9
SLAMMER-02 CU0 FC3 <-> SLAMMER-02 CU0 FS11 SLAMMER-02 CU1 FC3 <-> SLAMMER-02 CU1 FS10
***********
* WARNING *
***********
The following connection is invalid: SLAMMER-01 CU1 FC2 <-> SLAMMER-02 CU0 FS10
The following connection is invalid: SLAMMER-02 CU0 FC3 <-> SLAMMER-02 CU0 FS11
The following connection is invalid: SLAMMER-02 CU1 FC1 <-> SLAMMER-02 CU0 FS9
The following connection was not found: SLAMMER-01 CU1 FC2 <-> SLAMMER-02 CU0 FS9
The following connection was not found: SLAMMER-02 CU0 FC3 <-> SLAMMER-02 CU0 FS10
The following connection was not found: SLAMMER-02 CU1 FC1 <-> SLAMMER-02 CU0 FS8
|
