Skip to contentSkip to conversation

3 ultra100 controllers

DISCUSSION T. Ryan Halwachs 3 participants 6 messages Sep 4, 2002 – Sep 10, 2002

lore
Share & tools
lore
#0 lore raw

sent this to promise:

Quoted 7 lines

-----Original Message-----
From: T. Ryan Halwachs [mailto:[email protected]]
Sent: Monday, August 26, 2002 4:18 PM
To: [email protected]
Subject: 3 ultra100 controllers

Quoted 10 lines

hi,
i am trying to use three Promise technology controllers in one system.
the promise bios only shows drives attached to the first 2 controllers. how can i use the drives attached to the third controller?

cheers,
ryan

got this in reply:

Quoted 9 lines

From: support <[email protected]>
To: 'T. Ryan Halwachs' <[email protected]>
Subject: RE: 3 ultra100 controllers
Date: 28 Aug 2002 08:23:29 -0700

Hi Ryan
Sorry you will not be able to used 3 ultra100 in single system.
2 is the most.

Quoted 10 lines

It's All About Your Data!

Kevin Huynh
Reseller Technical Support.
Promise Technology, Inc
1745 McCandless Dr.
Milpitas Ca, 95035
408-228-6300

from pdc202xx.c in 2.4.19-ac4

/*
* linux/drivers/ide/pdc202xx.c Version 0.35 Mar. 30, 2002
*
* Copyright (C) 1998-2002 Andre Hedrick <[email protected]>
*

*  Promise Ultra33 cards with BIOS v1.20 through 1.28 will need this 
*  compiled into the kernel if you have more than one card installed. 
*  Note that BIOS v1.29 is reported to fix the problem.  Since this is 
*  safe chipset tuning, including this support is harmless 

*
* Promise Ultra66 cards with BIOS v1.11 this
* compiled into the kernel if you have more than one card installed. *
* Promise Ultra100 cards.
*

*  The latest chipset code will support the following :: 
*  Three Ultra33 controllers and 12 drives. 
*  8 are UDMA supported and 4 are limited to DMA mode 2 multi-word. 
*  The 8/4 ratio is a BIOS code limit by promise. 

*
* UNLESS you enable "CONFIG_PDC202XX_BURST"
*
*/

is it possible (using Linux) to run 3 (or more) PDC20267 based ultra100 cards in the same machine?

cheers,
ryan

#1 lore raw

BAWHAHAHA,

5 have been done!

Ask "Jeff Nguyen", all it means is that only two cards will be setup by their BIOS. The remaining cards will be setup by the driver. IIRC, there was a special RIO version with 8 card or 32 drives.

Cheers,

Quoted 68 lines · T. Ryan Halwachs

sent this to promise:

Quoted 7 lines

-----Original Message-----
From: T. Ryan Halwachs [mailto:[email protected]]
Sent: Monday, August 26, 2002 4:18 PM
To: [email protected]
Subject: 3 ultra100 controllers

Quoted 10 lines

hi,
i am trying to use three Promise technology controllers in one system.
the promise bios only shows drives attached to the first 2 controllers. how can i use the drives attached to the third controller?

cheers,
ryan

got this in reply:

Quoted 9 lines

From: support <[email protected]>
To: 'T. Ryan Halwachs' <[email protected]>
Subject: RE: 3 ultra100 controllers
Date: 28 Aug 2002 08:23:29 -0700

Hi Ryan
Sorry you will not be able to used 3 ultra100 in single system.
2 is the most.

Quoted 10 lines

It's All About Your Data!

Kevin Huynh
Reseller Technical Support.
Promise Technology, Inc
1745 McCandless Dr.
Milpitas Ca, 95035
408-228-6300

from pdc202xx.c in 2.4.19-ac4

/*
* linux/drivers/ide/pdc202xx.c Version 0.35 Mar. 30, 2002
*
* Copyright (C) 1998-2002 Andre Hedrick <[email protected]>
*

*  Promise Ultra33 cards with BIOS v1.20 through 1.28 will need this 
*  compiled into the kernel if you have more than one card installed. 
*  Note that BIOS v1.29 is reported to fix the problem.  Since this is 
*  safe chipset tuning, including this support is harmless 

*
* Promise Ultra66 cards with BIOS v1.11 this
* compiled into the kernel if you have more than one card installed. *
* Promise Ultra100 cards.
*

*  The latest chipset code will support the following :: 
*  Three Ultra33 controllers and 12 drives. 
*  8 are UDMA supported and 4 are limited to DMA mode 2 multi-word. 
*  The 8/4 ratio is a BIOS code limit by promise. 

*
* UNLESS you enable "CONFIG_PDC202XX_BURST"
*
*/

Quoted 5 lines

is it possible (using Linux) to run 3 (or more) PDC20267 based ultra100 cards in the same machine?

cheers,
ryan

Andre Hedrick
LAD Storage Consulting Group

#2 lore raw
Quoted 5 lines · Andre Hedrick

5 have been done!

Ask "Jeff Nguyen", all it means is that only two cards will be setup by their BIOS. The remaining cards will be setup by the driver. IIRC, there was a special RIO version with 8 card or 32 drives.

I wouldnt like to see the performance of the resulting box or try it on a VIA chipset or anything else I didn't trust 100% to handle contention on the PCI bus

#3 lore raw

Well Jeff will need to verify, but iirc many of those boxes he build and sold power "amazon.com". I got a charge knowing that factoid, when everyone was slamming the reliablity and strenghts of ATA servers.

Also there were select systems that may have been VIA but they were hand picked boards.

Anyways, Jeff will need to comment on this part. However I know he has shipped with 4 and maybe even 5 cards in a box that ran fine and stable.

Cheers,

Quoted 11 lines · Alan Cox
Quoted 5 lines · Andre Hedrick

5 have been done!

Ask "Jeff Nguyen", all it means is that only two cards will be setup by their BIOS. The remaining cards will be setup by the driver. IIRC, there was a special RIO version with 8 card or 32 drives.

I wouldnt like to see the performance of the resulting box or try it on a VIA chipset or anything else I didn't trust 100% to handle contention on the PCI bus

Andre Hedrick
LAD Storage Consulting Group

#4 lore raw
In reply to Alan Cox

kernel: 2.4.29-ac4
mobo: intel WS440BX (gateway OEM)
bios: 4W4SB0X0.15A.0019.P14
cpu: celeron 400
mem: 288MB
power: powmax LP6100 500W

controllers:
PDC20267(ultra100 PCI) PCI slot 1
PDC20267(ultra100 PCI) PCI slot 2
CMD649(ultraATA 100 PCI) PCI slot 3

drives:
1 WD1200AB 120G drive on card 1 primary channel
1 IC35L120AVVA07-0 120G drive on card 1 secondary channel 2 WD1200AB 120G drives on card 2 (each on a separate channel) 2 WD1200AB 120G drives on card 3 (each on a separate channel)

problem:
I cannot get these drives to work reliably in a software raid5 array. I was able to partition, mkraid, mke2fs, fsck, mount, and unmount.

I mounted the array locally on machine#0, then via nfs onto machine#1. I started copying files from machine#1's local drive(s) to the array on machine#0 via nfs.

I got these errors:

Sep  9 17:41:45 array kernel: hdi: dma_intr: status=0x51 { DriveReady
SeekComplete Error }
Sep  9 17:41:45 array kernel: hdi: dma_intr: error=0x84 {
DriveStatusError BadCRC }
Sep  9 17:42:12 array kernel: hdi: dma_intr: status=0x51 { DriveReady
SeekComplete Error }
Sep  9 17:42:12 array kernel: hdi: dma_intr: error=0x84 {
DriveStatusError BadCRC }
Sep  9 17:42:12 array kernel: hdi: dma_intr: status=0x51 { DriveReady
SeekComplete Error }
Sep  9 17:42:12 array kernel: hdi: dma_intr: error=0x84 {
DriveStatusError BadCRC }
Sep  9 17:42:16 array kernel: hdi: dma_intr: status=0x51 { DriveReady
SeekComplete Error }
Sep  9 17:42:16 array kernel: hdi: dma_intr: error=0x84 {
DriveStatusError BadCRC }
Sep  9 17:42:16 array kernel: hdi: dma_intr: status=0x51 { DriveReady
SeekComplete Error }
Sep  9 17:42:16 array kernel: hdi: dma_intr: error=0x84 {
DriveStatusError BadCRC }
Sep  9 17:42:16 array kernel: hdi: dma_intr: status=0x51 { DriveReady
SeekComplete Error }
Sep  9 17:42:16 array kernel: hdi: dma_intr: error=0x84 {
DriveStatusError BadCRC }
Sep  9 17:42:16 array kernel: hdi: dma_intr: status=0x51 { DriveReady
SeekComplete Error }
Sep  9 17:42:16 array kernel: hdi: dma_intr: error=0x84 {
DriveStatusError BadCRC }
Sep  9 17:42:16 array kernel: PDC202XX: Primary channel reset.
Sep  9 17:42:16 array kernel: ide4: reset: success
Sep  9 17:42:36 array kernel: hdk: timeout waiting for DMA
Sep  9 17:42:36 array kernel: PDC202XX: Secondary channel reset.
Sep  9 17:42:36 array kernel: hdk: ide_dma_timeout: Lets do it
again!stat = 0x50, dma_stat = 0x20
Sep  9 17:42:36 array kernel: hdk: DMA disabled
Sep  9 17:42:36 array kernel: PDC202XX: Secondary channel reset.
Sep  9 17:42:36 array kernel: hdk: ide_set_handler: handler not null;
old=c01c0840, new=c01bdf30
Sep  9 17:42:36 array kernel: bug: kernel timer added twice at c01c06a8.
Sep  9 17:42:36 array kernel: hdk: ide_set_handler: handler not null;
old=c01bdf30, new=c01bdfa0
Sep  9 17:42:36 array kernel: bug: kernel timer added twice at c01c06a8.
Sep  9 17:42:36 array kernel: hdi: timeout waiting for DMA
Sep  9 17:42:36 array kernel: PDC202XX: Primary channel reset.
Sep  9 17:42:36 array kernel: hdi: ide_dma_timeout: Lets do it
again!stat = 0x50, dma_stat = 0x20
Sep  9 17:42:36 array kernel: hdi: DMA disabled
Sep  9 17:42:36 array kernel: PDC202XX: Primary channel reset.
Sep  9 17:42:36 array kernel: hdi: ide_set_handler: handler not null;
old=c01c0840, new=c01bdf30
Sep  9 17:42:36 array kernel: bug: kernel timer added twice at c01c06a8.
Sep  9 17:42:36 array kernel: hdi: ide_set_handler: handler not null;
old=c01bdf30, new=c01bdfa0
Sep  9 17:42:36 array kernel: bug: kernel timer added twice at c01c06a8.
Sep  9 17:43:06 array kernel: hdk: dma_intr: status=0x58 { DriveReady
SeekComplete DataRequest }
Sep  9 17:43:06 array kernel: 
Sep  9 17:43:06 array kernel: hdk: status timeout: status=0xd0 { Busy }
Sep  9 17:43:06 array kernel: 
Sep  9 17:43:06 array kernel: hdk: DMA disabled
Sep  9 17:43:06 array kernel: PDC202XX: Secondary channel reset.
Sep  9 17:43:06 array kernel: hdk: drive not ready for command
Sep  9 17:43:06 array kernel: ide5: reset: success
Sep  9 17:43:26 array kernel: hdi: dma_intr: status=0x58 { DriveReady
SeekComplete DataRequest }
Sep  9 17:43:26 array kernel: 
Sep  9 17:43:26 array kernel: hdi: status timeout: status=0xd0 { Busy }
Sep  9 17:43:26 array kernel: 
Sep  9 17:43:26 array kernel: hdi: DMA disabled
Sep  9 17:43:26 array kernel: PDC202XX: Primary channel reset.
Sep  9 17:43:26 array kernel: hdi: drive not ready for command
Sep  9 17:43:26 array kernel: ide4: reset: success
Sep  9 18:07:17 array kernel: hdk: status timeout: status=0xd0 { Busy }
Sep  9 18:07:17 array kernel: 
Sep  9 18:07:17 array kernel: PDC202XX: Secondary channel reset.
Sep  9 18:07:17 array kernel: hdk: no DRQ after issuing WRITE
Sep  9 18:07:17 array kernel: ide5: reset: success
Sep  9 19:29:16 array kernel: hdi: status error: status=0x58 {
DriveReady SeekComplete DataRequest }
Sep  9 19:29:16 array kernel: 
Sep  9 19:29:16 array kernel: hdi: drive not ready for command
Sep  9 19:36:37 array kernel: hdi: status error: status=0x58 {
DriveReady SeekComplete DataRequest }
Sep  9 19:36:37 array kernel: 
Sep  9 19:36:37 array kernel: hdi: drive not ready for command
Sep  9 19:39:18 array kernel: hdi: status error: status=0x58 {
DriveReady SeekComplete DataRequest }
Sep  9 19:39:18 array kernel: 
Sep  9 19:39:18 array kernel: hdi: drive not ready for command
Sep  9 19:42:41 array kernel: hdi: status error: status=0x58 {
DriveReady SeekComplete DataRequest }
Sep  9 19:42:41 array kernel: 
Sep  9 19:42:41 array kernel: hdi: drive not ready for command
Sep  9 19:42:46 array kernel: hdi: status error: status=0x58 {
DriveReady SeekComplete DataRequest }
Sep  9 19:42:46 array kernel: 
Sep  9 19:42:46 array kernel: hdi: drive not ready for command
Sep  9 19:43:09 array kernel: hdi: status error: status=0x58 {
DriveReady SeekComplete DataRequest }
Sep  9 19:43:09 array kernel: 
Sep  9 19:43:09 array kernel: hdi: drive not ready for command
Sep  9 19:43:09 array kernel: hdi: status error: status=0x58 {
DriveReady SeekComplete DataRequest }
Sep  9 19:43:09 array kernel: 
Sep  9 19:43:09 array kernel: hdi: drive not ready for command
Sep  9 19:44:37 array kernel: hdi: status error: status=0x58 {
DriveReady SeekComplete DataRequest }
Sep  9 19:44:37 array kernel: 
Sep  9 19:44:37 array kernel: hdi: drive not ready for command
Sep  9 19:45:01 array kernel: hdi: status error: status=0x58 {
DriveReady SeekComplete DataRequest }
Sep  9 19:45:01 array kernel: 
Sep  9 19:45:01 array kernel: hdi: drive not ready for command
Sep  9 19:48:29 array kernel: hdi: status error: status=0x58 {
DriveReady SeekComplete DataRequest }
Sep  9 19:48:29 array kernel: 
Sep  9 19:48:29 array kernel: hdi: drive not ready for command
Sep  9 19:48:34 array kernel: hdi: status error: status=0x58 {
DriveReady SeekComplete DataRequest }
Sep  9 19:48:34 array kernel: 
Sep  9 19:48:34 array kernel: hdi: drive not ready for command
Sep  9 19:48:39 array kernel: hdk: status timeout: status=0xd0 { Busy }
Sep  9 19:48:39 array kernel: 
Sep  9 19:48:39 array kernel: PDC202XX: Secondary channel reset.
Sep  9 19:48:39 array kernel: hdk: no DRQ after issuing WRITE
Sep  9 19:48:39 array kernel: ide5: reset: success
Sep  9 19:55:05 array kernel: hdk: status timeout: status=0xd0 { Busy }
Sep  9 19:55:05 array kernel: 
Sep  9 19:55:05 array kernel: PDC202XX: Secondary channel reset.
Sep  9 19:55:05 array kernel: hdk: no DRQ after issuing WRITE
Sep  9 19:55:05 array kernel: ide5: reset: success
Sep  9 19:55:47 array kernel: hdi: status error: status=0x58 {
DriveReady SeekComplete DataRequest }
Sep  9 19:55:47 array kernel: 
Sep  9 19:55:47 array kernel: hdi: drive not ready for command
Sep  9 20:04:01 array kernel: hdi: status error: status=0x58 {
DriveReady SeekComplete DataRequest }
Sep  9 20:04:01 array kernel: 
Sep  9 20:04:01 array kernel: hdi: drive not ready for command
Sep  9 20:04:08 array kernel: hdi: status error: status=0x58 {
DriveReady SeekComplete DataRequest }
Sep  9 20:04:08 array kernel: 
Sep  9 20:04:08 array kernel: hdi: drive not ready for command
Sep  9 20:04:13 array kernel: hdi: status error: status=0x58 {
DriveReady SeekComplete DataRequest }
Sep  9 20:04:13 array kernel: 
Sep  9 20:04:13 array kernel: hdi: drive not ready for command
Sep  9 20:04:24 array kernel: hdi: status error: status=0x58 {
DriveReady SeekComplete DataRequest }
Sep  9 20:04:24 array kernel: 
Sep  9 20:04:24 array kernel: hdi: drive not ready for command
Sep  9 20:04:42 array kernel: hdi: status error: status=0x58 {
DriveReady SeekComplete DataRequest }
Sep  9 20:04:42 array kernel: 
Sep  9 20:04:42 array kernel: hdi: drive not ready for command
Sep  9 20:04:53 array kernel: hdi: status error: status=0x58 {
DriveReady SeekComplete DataRequest }
Sep  9 20:04:53 array kernel: 
Sep  9 20:04:53 array kernel: hdi: drive not ready for command
Sep  9 20:05:11 array kernel: hdi: status error: status=0x58 {
DriveReady SeekComplete DataRequest }
Sep  9 20:05:11 array kernel: 
Sep  9 20:05:11 array kernel: hdi: drive not ready for command
Sep  9 20:05:39 array kernel: hdi: status error: status=0x58 {
DriveReady SeekComplete DataRequest }
Sep  9 20:05:39 array kernel: 
Sep  9 20:05:39 array kernel: hdi: drive not ready for command
Sep  9 20:05:46 array kernel: hdi: status error: status=0x58 {
DriveReady SeekComplete DataRequest }
Sep  9 20:05:46 array kernel: 
Sep  9 20:05:46 array kernel: hdi: drive not ready for command
Sep  9 20:06:38 array kernel: hdi: status error: status=0x58 {
DriveReady SeekComplete DataRequest }
Sep  9 20:06:38 array kernel: 
Sep  9 20:06:38 array kernel: hdi: drive not ready for command
Sep  9 20:24:13 array kernel: hdk: status timeout: status=0xd0 { Busy }
Sep  9 20:24:13 array kernel: 
Sep  9 20:24:13 array kernel: PDC202XX: Secondary channel reset.
Sep  9 20:24:13 array kernel: hdk: no DRQ after issuing WRITE
Sep  9 20:24:14 array kernel: ide5: reset: success
Sep  9 20:26:28 array kernel: hdk: status timeout: status=0xd0 { Busy }
Sep  9 20:26:28 array kernel: 
Sep  9 20:26:28 array kernel: PDC202XX: Secondary channel reset.
Sep  9 20:26:28 array kernel: hdk: no DRQ after issuing WRITE
Sep  9 20:26:28 array kernel: ide5: reset: success
Sep  9 21:02:28 array kernel: hdk: status timeout: status=0xd0 { Busy }
Sep  9 21:02:28 array kernel: 
Sep  9 21:02:28 array kernel: PDC202XX: Secondary channel reset.
Sep  9 21:02:28 array kernel: hdk: no DRQ after issuing WRITE
Sep  9 21:02:28 array kernel: ide5: reset: success
Sep  9 21:04:58 array kernel: hdk: status timeout: status=0xd0 { Busy }
Sep  9 21:04:58 array kernel: 
Sep  9 21:04:58 array kernel: PDC202XX: Secondary channel reset.
Sep  9 21:04:58 array kernel: hdk: no DRQ after issuing WRITE
Sep  9 21:04:58 array kernel: ide5: reset: success
Sep  9 21:55:05 array kernel: hdk: status timeout: status=0xd0 { Busy }
Sep  9 21:55:05 array kernel: 
Sep  9 21:55:05 array kernel: PDC202XX: Secondary channel reset.
Sep  9 21:55:05 array kernel: hdk: no DRQ after issuing WRITE
Sep  9 21:55:05 array kernel: ide5: reset: success

now machine#0 hangs at various places during boot.
I had tried other -ac kernels as well as some 2.5.xx. No joy. Is there any work done lately that may address this(these) problems?

thanks a whole lot,
ryan

#5 lore raw
Quoted 1 line · T. Ryan Halwachs

kernel: 2.4.29-ac4

Should be 2.4.19-ac4