RAID with HP P800

[yan1982]

Hi,

We have a problem with a HP Smart Array P800 with 25 HDD, ext3 partition.

We have found (we think):
- order HDD (1 to 25)
- block size (16 kb)
- RAID 6 (ADG)
- Parity left async.
- stripe size jump (16)

We have on the first hdd:
- sector 0 => the MBR
- sector 1 => EFI PART
- sector 2 => 1 descriptor table, start 0x22 => 34

Problem:
- we can't view ext3 partition

We suppose, there is a offset on HDD 2-25 but how many but 0 offset on HDD 1 ? It's possible ?

Any idea are welcome.

Bye

[ddd123]

try to create new 4 HDDs raid for test. and look on the data

[yan1982]

Yes I want... but I have only HDD... not the bay....

[DR-Kiev]

Hi,

We have a problem with a HP Smart Array P800 with 25 HDD, ext3 partition.

We have found (we think):
- order HDD (1 to 25)
- block size (16 kb)
- RAID 6 (ADG)
- Parity left async.
- stripe size jump (16)

We have on the first hdd:
- sector 0 => the MBR
- sector 1 => EFI PART
- sector 2 => 1 descriptor table, start 0x22 => 34

Problem:
- we can't view ext3 partition

We suppose, there is a offset on HDD 2-25 but how many but 0 offset on HDD 1 ? It's possible ?

Any idea are welcome.

Bye

Nice one

I guess you were trying to figure out it in UFS Explorer? Right?
As i know it couldn't build Raid6 with delays .

You have only chance to figure out it in R-studio
But , it is don't know correct Raid6 configurations for HP
So, only manual work in R-studio .
But... your array will have ....25*16=400 sequences
and 9200 cells

[yan1982]

Yes, it's true

Do you want the XML... we have build a software for build this xml...

[DR-Kiev]

Yes, it's true
Do you want the XML... .

No thanks, i have it

we have build a software for build this xml...

I don't get , you have built or still building ?

[yan1982]

We have made a software for generate this XML file.

[DR-Kiev]

We have made a software for generate this XML file.

Great , nice job

Show us your Matrix for admire

[yan1982]

Thanks....

First 166 rows...

But I you have an solution, it's better....

Attachments

raid_xml.zip

(13.69 KiB) Downloaded 582 times

[DR-Kiev]

Thanks....

First 166 rows...

But I you have an solution, it's better....

Almost fine except one mistake (will not work if you will be missing one or two drives) .
And show us the last 16 sequences, there is some trick too.

[yan1982]

!!!!!!

You can find the schema... We can't view some trick....

<Block id="A399">ReedSolomon</Block>
<Block id="B399">9155</Block>
<Block id="C399">9156</Block>
<Block id="D399">9157</Block>
<Block id="E399">9158</Block>
<Block id="F399">9159</Block>
<Block id="G399">9160</Block>
<Block id="H399">9161</Block>
<Block id="I399">9162</Block>
<Block id="J399">9163</Block>
<Block id="K399">9164</Block>
<Block id="L399">9165</Block>
<Block id="M399">9166</Block>
<Block id="N399">9167</Block>
<Block id="O399">9168</Block>
<Block id="P399">9169</Block>
<Block id="Q399">9170</Block>
<Block id="R399">9171</Block>
<Block id="S399">9172</Block>
<Block id="T399">9173</Block>
<Block id="U399">9174</Block>
<Block id="V399">9175</Block>
<Block id="W399">9176</Block>
<Block id="X399">9177</Block>
<Block id="Y399">XorOfData</Block>
<Block id="A400">ReedSolomon</Block>
<Block id="B400">9178</Block>
<Block id="C400">9179</Block>
<Block id="D400">9180</Block>
<Block id="E400">9181</Block>
<Block id="F400">9182</Block>
<Block id="G400">9183</Block>
<Block id="H400">9184</Block>
<Block id="I400">9185</Block>
<Block id="J400">9186</Block>
<Block id="K400">9187</Block>
<Block id="L400">9188</Block>
<Block id="M400">9189</Block>
<Block id="N400">9190</Block>
<Block id="O400">9191</Block>
<Block id="P400">9192</Block>
<Block id="Q400">9193</Block>
<Block id="R400">9194</Block>
<Block id="S400">9195</Block>
<Block id="T400">9196</Block>
<Block id="U400">9197</Block>
<Block id="V400">9198</Block>
<Block id="W400">9199</Block>
<Block id="X400">9200</Block>
<Block id="Y400">XorOfData</Block>
</Table>
</RAID>
</RAIDList>

[DR-Kiev]

Looks good. End right.
Have you found offsets already?
Usual 1088
And, as i hinted you before , interchange XOR and Reed-Solomon (will help if you have missed disks).