In my haste for help, I stupidly hit reply and changed the subject which I thought was enough for a new message, not giving much thought for the threading, etc. So apologies for the hijack, although I would think it fairly obvious that it wasn't deliberate or obvious to me... or a big deal really. My mail doesn't thread.
Some further info for those actually interested in helping, I forgot that the HD in question had an Ubuntu 8.10 install on it, so I fired that up and hdparm -t produced 56MB/Sec. So the issue is either kernel or config.
Thanks for your patience.
----- Forwarded Message ----
I have a system who's normal activity is running Xen 3.3.1 with Centos 5.3 as the DomU. I had reason to connect an IDE drive to the single IDE interface that this board has. I was getting slow performance. hdparm -t revealed that the thoughput was down to just 2.7 MB/sec. After some tweaking (switch on 32 bit,etc) and retesting under a non-Xen config, I managed to get that to 8MB/Sec. When I move the drive to a USB to IDE adapter, hdparm -t gives me about 18MB/Sec. Big improvement, although I don't know if it is then perfect..
My kernel is 2.6.18-128.4.1.el5. I have kernel parameter pci=nomsi. I can't remember exactly why I needed this, but there was a good reason, maybe SATA drive not working properly or some such. Could this be the cause of my issue?
The motherboard blurb states that the IDE interface is ATA/133 compliant and I am using the provided IDE cable. I have tried three hard disks, all with similar results. (2x 40gb + 1x 160gb)
lspci | grep IDE gives 00:06.0 IDE interface: nVidia Corporation MCP78S [GeForce 8200] IDE (rev a1)
As a reference, hdparm -t against the SATA drive in AHCI mode on the same board gives about 110MB/Sec.
Any ideas, anyone?
Thanks in advance,
Ian.
2009/8/11 Ian Murray murrayie@yahoo.co.uk:
In my haste for help, I stupidly hit reply and changed the subject which I thought was enough for a new message, not giving much thought for the threading, etc. So apologies for the hijack, although I would think it fairly obvious that it wasn't deliberate or obvious to me... or a big deal really. My mail doesn't thread.
Some further info for those actually interested in helping, I forgot that the HD in question had an Ubuntu 8.10 install on it, so I fired that up and hdparm -t produced 56MB/Sec. So the issue is either kernel or config.
Check if ultradma and unmasq interrupts flags are set(both under ubuntu and centos):
hdparm -u -d /dev/hda
Hi,
Thanks for the response.
Under Ubuntu (NB the drive was under /dev/sdb under Ubuntu - is this a clue?!?):-
HDIO_GET_UNMASKINTR failed: Inappropriate ioctl for device HDIO_GET_DMA failed: Inappropriate ioctl for device
Under Centos (as /dev/hda):-
Both set to 0, so I tried to set them to 1. -d1 error as inappropriate but -u was allowed to be set. It didn't make much difference to the performance, i.e. up from ~ 2.5MB/Sec to 4MB/Sec.
as a refernence, -I gives...
ATA device, with non-removable media Model Number: Maxtor 6E040L0 Serial Number: E1L01XLE Firmware Revision: NAR61590 Standards: Used: ATA/ATAPI-7 T13 1532D revision 0 Supported: 7 6 5 4 Configuration: Logical max current cylinders 16383 16383 heads 16 16 sectors/track 63 63 -- CHS current addressable sectors: 16514064 LBA user addressable sectors: 78156288 device size with M = 1024*1024: 38162 MBytes device size with M = 1000*1000: 40016 MBytes (40 GB) Capabilities: LBA, IORDY(can be disabled) Standby timer values: spec'd by Standard, no device specific minimum R/W multiple sector transfer: Max = 16 Current = 16 Advanced power management level: unknown setting (0x00fe) Recommended acoustic management value: 192, current value: 0 DMA: mdma0 mdma1 mdma2 udma0 udma1 udma2 udma3 udma4 udma5 *udma6 Cycle time: min=120ns recommended=120ns PIO: pio0 pio1 pio2 pio3 pio4 Cycle time: no flow control=120ns IORDY flow control=120ns Commands/features: Enabled Supported: * SMART feature set Security Mode feature set * Power Management feature set * Write cache * Look-ahead * Host Protected Area feature set * WRITE_VERIFY command * WRITE_BUFFER command * READ_BUFFER command * NOP cmd * DOWNLOAD_MICROCODE * Advanced Power Management feature set SET_MAX security extension Automatic Acoustic Management feature set * Device Configuration Overlay feature set * Mandatory FLUSH_CACHE * FLUSH_CACHE_EXT * SMART error logging * SMART self-test Security: Master password revision code = 65534 supported not enabled not locked frozen not expired: security count not supported: enhanced erase HW reset results: CBLID- above Vih Device num = 0 determined by CSEL Checksum: correct
Thanks,
Ian.
________________________________ From: Alexander Georgiev alexander.georgiev@gmail.com To: CentOS mailing list centos@centos.org Sent: Tuesday, 11 August, 2009 8:57:17 Subject: Re: [CentOS] Slow IDE on GeForce 8200 board
2009/8/11 Ian Murray murrayie@yahoo.co.uk:
In my haste for help, I stupidly hit reply and changed the subject which I thought was enough for a new message, not giving much thought for the threading, etc. So apologies for the hijack, although I would think it fairly obvious that it wasn't deliberate or obvious to me... or a big deal really. My mail doesn't thread.
Some further info for those actually interested in helping, I forgot that the HD in question had an Ubuntu 8.10 install on it, so I fired that up and hdparm -t produced 56MB/Sec. So the issue is either kernel or config.
Check if ultradma and unmasq interrupts flags are set(both under ubuntu and centos):
hdparm -u -d /dev/hda _______________________________________________ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
On 08/12/2009 01:34 AM, Ian Murray wrote: ...
Under Ubuntu (NB the drive was under /dev/sdb under Ubuntu - is this a clue?!?):-
Some HP machines have the same symptoms you describe:
Slow disk IO with CentOS, fast with Fedora.
The cure is to append
hda=noprobe hdc=noprobe
to the startup parameters in /boot/grub/grub.conf.
Mogens
Thanks for the response. Unfortunately /dev/hda was unavailable after I tried that.
________________________________ From: Mogens Kjaer mk@crc.dk To: CentOS mailing list centos@centos.org Sent: Wednesday, 12 August, 2009 6:29:16 Subject: Re: [CentOS] Slow IDE on GeForce 8200 board
On 08/12/2009 01:34 AM, Ian Murray wrote: ...
Under Ubuntu (NB the drive was under /dev/sdb under Ubuntu - is this a clue?!?):-
Some HP machines have the same symptoms you describe:
Slow disk IO with CentOS, fast with Fedora.
The cure is to append
hda=noprobe hdc=noprobe
to the startup parameters in /boot/grub/grub.conf.
Mogens
On 08/12/2009 09:38 AM, Ian Murray wrote:
Thanks for the response. Unfortunately /dev/hda was unavailable after I tried that.
The disks on my HP machines show up as /dev/sda when I boot with hda=noprobe hdc=noprobe.
Mogens
-
Alexander Georgiev -
Ian Murray -
Mogens Kjaer