PERC 5/i RAID Card: Tips and Benchmarks - Page 137 - Overclock.net

Forum Jump: 
Reply
 
Thread Tools
post #1361 of 7276 Old 03-03-2009, 03:48 PM
4.0ghz
 
bruestle2's Avatar
 
Join Date: Apr 2007
Location: Pennsylvania, USA
Posts: 2,016
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 118
Given that they did MAKE the card, there is a good chance they will. This still falls under what I was talking about before where they may not because they want as little liability as possible. But the fact that they manufacture the card means they will probably support it anyway.

Quote:
Originally Posted by t4ct1c47
No need to break out fancy chemicals when you have bodily fluids.

Chimp Challenge Participant  7 Million+ Folding at Home points 

bruestle2 is offline  
Sponsored Links
Advertisement
 
post #1362 of 7276 Old 03-03-2009, 03:56 PM
Commodore 64
 
bbarnett's Avatar
 
Join Date: Jan 2009
Posts: 19
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 2
Quote:
Originally Posted by bruestle2 View Post
Given that they did MAKE the card, there is a good chance they will. This still falls under what I was talking about before where they may not because they want as little liability as possible. But the fact that they manufacture the card means they will probably support it anyway.
Oh geez.

Sure, they support THE CARD. The HARDWARE... although their customer is DELL, not you! This bug, however, is with the SOFTWARE on the card, and no.. LSI is not going to support Dell's software.

Why on earth would they? Why would LSI modify their firmware to work with another piece of software, that they have nothing to do with?! Modifying their firmware won't help Dell, or Dell customers! It won't help LSI, or LSI's customers.

The only people in the entire world it will help, is a bunch of people on a few boards, most of which don't EVEN NEED TO RUN THIS FIRMWARE TO USE THE CARD.

I mean, good grief, what are you thinking?? THINK, MAN! THINK ABOUT IT.
bbarnett is offline  
post #1363 of 7276 Old 03-03-2009, 04:16 PM
 
iakovl's Avatar
 
Join Date: Jan 2009
Location: Israel
Posts: 281
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 4
so... after all of this ^. i gonna use R5 with 4 F1 1Tera hdd's in xp, am i safe of fu*ked ?
i use LSI FW 7.0.1-0064 right now

‿‿ ◕ ‿‿ ◕ ‿‿ ◕ ‿‿ ◕ ‿‿
iakovl is offline  
Sponsored Links
Advertisement
 
post #1364 of 7276 Old 03-03-2009, 04:21 PM
Commodore 64
 
bbarnett's Avatar
 
Join Date: Jan 2009
Posts: 19
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 2
Quote:
Originally Posted by iakovl View Post
so... after all of this ^. i gonna use R5 with 4 F1 1Tera hdd's in xp, am i safe of fu*ked ?
i use LSI FW 7.0.1-0064 right now
The truth of the matter is this:

1) there is a known bug already
2) there is no testing done by anyone, save for the people here, for this config
3) the people here are a very small number of people, and most are not qualified engineers, and are not performing rigorous testing in a controlled environment

The question you should be asking yourself is this. Do you care about your data?

If so, then ask this next question. Why would you go from a tested and hardened config, to a known to be buggy config? Why do you need that LSI Logic firmware?
bbarnett is offline  
post #1365 of 7276 Old 03-03-2009, 05:12 PM
4.0ghz
 
bruestle2's Avatar
 
Join Date: Apr 2007
Location: Pennsylvania, USA
Posts: 2,016
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 118
Quote:
Originally Posted by bbarnett View Post
Oh geez.

Sure, they support THE CARD. The HARDWARE... although their customer is DELL, not you! This bug, however, is with the SOFTWARE on the card, and no.. LSI is not going to support Dell's software.

Why on earth would they? Why would LSI modify their firmware to work with another piece of software, that they have nothing to do with?! Modifying their firmware won't help Dell, or Dell customers! It won't help LSI, or LSI's customers.

The only people in the entire world it will help, is a bunch of people on a few boards, most of which don't EVEN NEED TO RUN THIS FIRMWARE TO USE THE CARD.

I mean, good grief, what are you thinking?? THINK, MAN! THINK ABOUT IT.
Maybe because they care about their customers. The fact that we bought this product second hand still makes us one of their customers indirectly. LSI sold the card to Dell, Dell sold the card to somebody, and that somebody sold the card to us. This is how almost all sales work. Any time that you buy something from walmart, newegg, or ANY retalier, this is how it works. The original manufacturer supports their indirect customers and the retailer supports their direct customers.

Anyway, I'm not going to continue this conversation with you. I can see now this is turning into a full blown argument. If you don't want to call LSI, that's fine. But for the sake of everyone here who you are trying to help, I wish you would. I would do it myself, but I can't call LSI about a problem I am not experiencing. LSI should be able to clarify if this is an issue with their firmware or with the Dell boot loader. I do not expect them to support the boot loader, though I would expect them to support their firmware (assuming it is only an issue with the firmware).

Thank you for the heads up though!

Bruestle II

Quote:
Originally Posted by t4ct1c47
No need to break out fancy chemicals when you have bodily fluids.

Chimp Challenge Participant  7 Million+ Folding at Home points 

bruestle2 is offline  
post #1366 of 7276 Old 03-03-2009, 05:33 PM
3.0ghz
 
devsk's Avatar
 
Join Date: Jan 2009
Posts: 523
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 27
One of the issues of the card I found was that Dell firmware was severely crippled in terms of performance. I got about 20-25% better throughput using LSI firmware. So, I do have a reason for using LSI firmware.

Now, coming to the bug: Any piece of software can have bugs. Its a LSI problem. And I think there is a very good chance that they would want to fix it, even if the card has Dell's bootloader. If you think Dell's firmware is a complete rewrite or something totally different from LSI, you are wrong. And this is coming from looking at the source of their drivers for linux. Next time they sync their code base, they are going to see this bug anyway.

One thing you did not mention is which LSI firmware you were using? Did you try a different firmware from LSI e.g. an older or newer version than the one which gave you the issue?
devsk is offline  
post #1367 of 7276 Old 03-03-2009, 05:51 PM
3.0ghz
 
devsk's Avatar
 
Join Date: Jan 2009
Posts: 523
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 27
Quote:
Originally Posted by bruestle2 View Post
Anyway, I'm not going to continue this conversation with you. I can see now this is turning into a full blown argument.
Problem with this guy is that he is patronizing in attitude. Ok, agreed you hit a problem with a recommended setup but you just can't keep looking down on other people just because they happen to use something which is considered i33t by you.

Barnett still did not describe in detail what exactly happened? He was using the card, and suddenly one day he ran a hexdump on a block of a device on the RAID and found that the hexdump did not match with the one inside his head and he exclaimed BUG! What were the real symptoms?

He also did not specify which LSI firmware version is affected?

I have blocks towards the end of my devices which contain RAID signatures (part of DDF standard to identify RAID devices) which contain strings identifying Dell Perc 5/i as my RAID controller.

I did not see the command which created that hexdump that he posted.
devsk is offline  
post #1368 of 7276 Old 03-03-2009, 05:52 PM
Commodore 64
 
bbarnett's Avatar
 
Join Date: Jan 2009
Posts: 19
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 2
Quote:
Originally Posted by bruestle2 View Post
Maybe because they care about their customers. The fact that we bought this product second hand still makes us one of their customers indirectly. LSI sold the card to Dell, Dell sold the card to somebody, and that somebody sold the card to us. This is how almost all sales work. Any time that you buy something from walmart, newegg, or ANY retalier, this is how it works. The original manufacturer supports their indirect customers and the retailer supports their direct customers.
This is not the same thing, what so ever.

You aren't buying a card that states "LSI" anywhere on it. You aren't buying a card that is sold as an LSI card. You're not going to Dell and adding an LSI card to the skew for your server...

This is not the same as a retailer selling a product to you.

What this is, is someone buying a part (the hardware) from LSI, and licensing the software from them. Then, they are modifying that software in ways that makes it incompatible with the original part.

Dell has its own raid manager, its own firmware, its own boot loader, everything from the ground up.

A close approximation would be if you bought a Pontiac Vibe, and then called Toyota and demanded warranty support because you flashed the Matrix's software for the ECU, onto the Vibe, and now the ABS didn't work. You'd be laughed out of the dealership!

Quote:
Anyway, I'm not going to continue this conversation with you. I can see now this is turning into a full blown argument. If you don't want to call LSI, that's fine. But for the sake of everyone here who you are trying to help, I wish you would. I would do it myself, but I can't call LSI about a problem I am not experiencing. LSI should be able to clarify if this is an issue with their firmware or with the Dell boot loader. I do not expect them to support the boot loader, though I would expect them to support their firmware (assuming it is only an issue with the firmware).

Thank you for the heads up though!

Bruestle II
Dude,

You've been told where this issue occurs. It is an issue with the Dell bootloader, and LSI firmware, not being compatible. "Supporting" their firmware could *only* happen if they make it compatible with the Dell bootloader. That would mean:

1) making it incompatible with their bootloader

or

2) increasing their codebase size, so it supports both the Dell and LSI bootloader.

Why would they do either of these things?

You seem to think that LSI should work to fix that problem. There isn't one though. There is no bug here, except you are using the wrong firmware.
bbarnett is offline  
post #1369 of 7276 Old 03-03-2009, 06:02 PM
Commodore 64
 
bbarnett's Avatar
 
Join Date: Jan 2009
Posts: 19
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 2
Quote:
Originally Posted by devsk View Post
One of the issues of the card I found was that Dell firmware was severely crippled in terms of performance. I got about 20-25% better throughput using LSI firmware. So, I do have a reason for using LSI firmware.
I would email Dell with that, then... instead of trying non-verified code on a raid card which you (I imagine) have data you care about.

Quote:
Now, coming to the bug: Any piece of software can have bugs. Its a LSI problem.
No, it is not.

Quote:
And I think there is a very good chance that they would want to fix it, even if the card has Dell's bootloader. If you think Dell's firmware is a complete rewrite or something totally different from LSI, you are wrong. And this is coming from looking at the source of their drivers for linux. Next time they sync their code base, they are going to see this bug anyway.
Actually, you could have the same driver in Linux, supporting two entirely different company's hardware products, with entirely different codebases. Only the API would need to be the same, and we're not talking about an OS/ hardware bug here.

As well, there is no bug in the LSI firmware.

This is not a piece of code, where you have to sanatize things so that no matter what user input is provided, you don't have a buffer overflow. This is a piece of software that communicates only with another piece of software, during a certain part of the boot process.

Quote:
One thing you did not mention is which LSI firmware you were using? Did you try a different firmware from LSI e.g. an older or newer version than the one which gave you the issue?
No, that would be pointless.
bbarnett is offline  
post #1370 of 7276 Old 03-03-2009, 06:09 PM
Commodore 64
 
bbarnett's Avatar
 
Join Date: Jan 2009
Posts: 19
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 2
Quote:
Originally Posted by devsk View Post
Problem with this guy is that he is patronizing in attitude. Ok, agreed you hit a problem with a recommended setup but you just can't keep looking down on other people just because they happen to use something which is considered i33t by you.

Barnett still did not describe in detail what exactly happened? He was using the card, and suddenly one day he ran a hexdump on a block of a device on the RAID and found that the hexdump did not match with the one inside his head and he exclaimed BUG! What were the real symptoms?

He also did not specify which LSI firmware version is affected?

I have blocks towards the end of my devices which contain RAID signatures (part of DDF standard to identify RAID devices) which contain strings identifying Dell Perc 5/i as my RAID controller.

I did not see the command which created that hexdump that he posted.
Well, you've missed some of my posts.

Naturally, I did not simply run around my raid card, and do hex dumps at random.. then see some bad data, and think "BUG!". Don't be silly.

Perhaps you would like to be more reasonable in your assertions?
bbarnett is offline  
Reply

Quick Reply

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off