Hi Justin,

I used to work many many years ago in Sun support level 2, mainly doing NBU support. (Have never worked for Oracle, if that helps date my time with Sun).

So, next I should say that this is my opinion/experience and not that of Sun’s and not a reflection on any one individual.

But I can quickly give some insight/tips:

-          Many but not all customers see the request to update firmware as stalling. Don’t dispute that this is a common opinion.

-          The engineers are instructed to do this – they have no choice – so don’t get too grumpy at them. This is one of many due diligence things they are expected to do – it only stands out because it involves an outage.

-          Some don’t like it because they only want to make technically justified changes – I agree with this, but in moderation/ in context. I eventually came to the position that 1. systems should be resourced/sufficiently redundant to allow an outage and 2. A firmware update is not like the support engineer is asking for an outage to the business’s principal database or a customer facing system, so it’s not such an diabolic request and needs to stay in perspective.

-          Technically, I see more than one reason to update the firmware.

o   The obvious reason is it’s a problem identified as fixed in the README.

o   The other one people generally think of is an issue might be accidentally fixed by the update.

o   But the one that isn’t so well-known is that level 3 support have diagnostic tools – it’s actually possible to collect logs from the tapedrive and examine these – the update to firmware source code doesn’t just improve functionality and remove code defects, but also improves log/debug information. It’s presumably still the case. I only found out about this tool(s) as I had a very competent dedicated level 3 engineer two cubicle walls over from me – it wasn’t particularly common knowledge or commonly needed.

Now to help you with your issue, my recommendation is:

1.       Schedule to do the update.

2.       Whilst you waiting for a window to do the update, run some vmstat and iostat commands collecting output – particularly monitor %b – there’s so much to interpret just with this, and you’re in a better position than an offsite support engineer to interpret it as you will know when the tape drive is and isn’t in use. This can identify a dying tape drive without looking at the diagnostic logs. (I’ve often wished customers would do this before ringing support as it’s the obvious thing to do before updating firmware) (I’m assuming your media server is on a nix box as you’ve mentioned Oracle support.)

3.       If iostat doesn’t show an issue and the problem persists after update, get the problem escalated until you have a level2/3 engineer who can look at diagnostics.

Hope this helps.

Robyn

Hi Len et all,

I looked in the error warn info logs and I could not find anything interesting, I have seen temp alarms in one site with SL500s, once, but not at this particular site.

Thanks for the information though, it seems > B5BF is highly recommended (Oracle also recommended moving to the newer version).

Justin.

Justin,

Do you have a library that will show you a little more detail on the error from the log that is within the tape drive and or the tape library. Also the code that shows up on the led display on the front of the tape drive. If any.

We have been having issues with some lto-5 tape drives that appears to be a cooling fan issue.

This shows up as a 3584 library issues.

HEC/HECQ B886 (Canister cooling fan failure)

We had  upgraded the firmware to address several other issue. We are now at B6W8.

Hi,

I read on this list awhile back there were some media issues others were having with the IBM LTO-5 tape drives; after an update, things were better.

Currently running B5BF and notice a lot of ‘(86) - media read errors’ on separate drives in different locations, has anyone seen this, one could chalk it up to some bad tape media but because it seems to occur across >2 drives in >1 location I was curious if anyone else had seen this issue and if so, which F/W where they running that seemed to solve the issue?

Drive: IBM LTO-5 8.0gbps F/C

F/W: B5BF

Justin.

No virus found in this message.
Checked by AVG - www.avg.com
Version: 10.0.1415 / Virus Database: 2102/4079 - Release Date: 12/13/11