MSA1000 - error 24 - BAD SCSI BUS MODE NON-LVD

[Martin Kerr]

Hi All,

wonder if someone can help... during a disk array re-organisation this week, we encountered this error...

the error seems incorrect:

the error arose 1/2 an hour after a hardware change [ addition of a 4314 box to the second SCSI channel i/f at the back ].

this change was to move 3 disks from a 6 disk lun into this other box for resilience across multiple scsi channels...

during this move 1 of the disks failed.

we have since tested the box, and the cable used on a test MSA system and all seem fine.

[ including the failed disk ]

the problem we now have is that we have a reported error which we need to identify the reason... as no 'non-ldv' devices had ever been connected.... the MSA seems to be fine since we undid the change - and removed the box. and the box and cable seem fine on another MSA...

we need to perform this upgrage again [ hopefully sucessfully ] next monday night, and we want to have elimited the problem before then.... and this is a live system with no downtime availability before then either... so we cannot take the MSA off line before the scheduled change...

... now, perhaps the SCSI interface is at fault, but 2 of the three disks were functioning fine.... and we really don't want to replace the whole MSA1000 when it otherwise seems quite happy [ it already has a box hanging off it's first external scsi port ]


any help on 'other reasons this error may occur' would be great...

thanks in advance,

Martin.

[Bostjan Kosi]

Hi,

maybe a defective backplain or SCSI I/O module in the 4314 box was the reason....
Or it was just one of those days....

[Martin Kerr]

:-)

yeah, i'd like to think it was just one of those days... however this is in a 24/7 critical system for government and i don't want next monday to also be one of those days... since we've got a downtime window of 5 hours, and 5 hours work to complete in that window....

in your opinion therefore, could you confirm there are no known alternative reasons for this error as far as the MSA1000 is concerned?

cheers.

[Bostjan Kosi]

I can not find any advisory stating any kind of this errors you are seeing so....

[John Kufrovich]

Martin,
By any chance, are you performing the LUN split will HOT/Live.

[John Kufrovich · Accepted solution]

With the enclosure attached, using cli, can you do a show versions. I would like to see what rev the FW is at.

[Martin Kerr]

Hi,

a) ... the MSA was powered down during the move of the three disks from the internal array enlosure out to the external 4314 box

b) the firmware on the MSA is 4.32 [ the latest I beleive ]


does that help ??? :-)

cheers,

Martin.

[John Kufrovich]

Really need the enclosure FW data.

If you have the cli cable, you can do a show version(s). This will dump all fw of everything attached to the MSA.

[Martin Kerr]

okay,

i will check later this afternoon... hp support have advised us to swap out the SCSI EMU Module from the MSA1000 for a replacement...

with regard to the firmware: this enclosure is working fine on another [test] MSA1000 - which has the same MSA FW levels as the live unit... but, as you say - something to check.

thanks again....

[John Kufrovich]

Didn't realize you had a test MSA. Then you can easily debug the problem. Swap MSA EMUs and see if the problem follows the bad EMU.

[Martin Kerr]

... couldn't swap for the test MSA .... the live MSA is a 24/7 email system covering the whole of my government here and I need things signed in triplicate before I can breathe at the live MSA/SAN

.... on the other hand... after your advice yesterday, i upgraded the test MSA f/w to enable it to report the external box fw [ MSA f/w 2.38 doesn't report it ], and all was correct anyhow.... i double checked this with the box fw on the live system [ I am allowed to use the CLI as long as i'm non destructive :-) ] ... and the box already plugged in to that MSA was the same levels as the one that we were going to add.... so nothing weird there...

HP flew over a spare MSA SCSI I/O just to be on the safe side, but the government told me i could not change it [ as it was not in the already agreed plan ]

so, last night, we went ahead with the scheduled down time, installed the box that failed last week, with a shorter cable, and all was and is okay....

so i'm more inclined to accept the first answer to this thread "it was one of those days"...

either way... panic over :-)

cheers for all your replies... i'll leave this thread open for another day, just in case theres any final comments that anyone thinks are fit to add...

thanks again,

martin.

[Martin Kerr]

... just to inform that this fault seems to have cleared with the only change being the SCSI cable used... [ although the original cable passed on a test system ]

thanks for all the input.

martin.

[Martin Kerr]

as above...