CentOS + itanium 2 + EFI = ARGH

[jkmccarthy]

I'm not sure how many of these were sold on ebay (I think the vendor had 20 or so up) but we can all post our revs and see if everyone's got the same.

Here's the details on my machine ...

American Megatrends Enterprise64
BIOS 12/08/2003 10:06:38

... (then successfully completes POST) ...

LSI Logic Corp, MPT BIOS

MPT-BIOS 5.07.00
... (then successfully scans SCSI bus and identifies drives) ...

EFI version 1.10 [14.61]
EFI IA-64 SDV/FDK (BIOS Callbacks) [Mon Dec 8 10:07:07 2003]
This image MainEntry is at 000000007FA02000

Copyright (c) 2000-2003 Broadcom Corporation
Broadcom NetXtreme Gigabit Ethernet EFI driver v6.1.2


_ <== more often than not, hangs with cursor here


I have a SIIG "active" PS2 to USB adaptor that connects PS2 keyboard and PS2 mouse into one of the four USB inputs in the back of the server. Based on pwfig's experiences, I'm now running with all three ethernet cables tied into an ethernet switch, but that has not cured my problems -- machine still gets hung after the NetXtreme Gigabit Ethernet report banner as shown above. Strange, since when it succeeds, the very next report banner is from the Broadcom NetXtreme Gigabit Ethernet "BCM" driver (or some such ... don't have details in front of me, since machine is currently being uncooperative, and hanging itself between those steps pretty regularly right now...).

Could hanging between the first NetXtreme message and the second NetXtreme message really be because of the SIIG PS2-to-USB keyboard+mouse not being sufficiently plain-vanilla and/or generic ? Stranger things have been known to occur, I'm sure ... I'll try this fix later in the week then.

-- Jim

P.S. I don't have convenient access to a plain-vanilla generic USB keyboard now, BUT what I did instead was disconnect the SIIG PS2-to-USB adaptor (basically unplug my keyboard and mouse), and as Securix suspected, this _DID_ fix the problem ... the machine no longer has a problem getting hung at that step ! (For now, I found if I plug in the PS2-to-USB active adaptor as soon as I see the second Broadcom NetXtreme banner report, the mouse+keyboard don't work in time for the EFI boot options menu, but by the time the [default boot option] Linux OS starts, the mouse+keyboard are OK and I'm able to login. Not a long-term solution [guess I'll be shopping for a bargain basement USB keyboard], but fine "for now" ....).

Many thanks to Securix for sharing his experience and suspecting my problem might be similar ... saved me from a much more lengthy troubleshooting exercise (e.g., re-init the BIOS using the motherboard jumper, etc. etc.)

 

[Securix]

Originally posted by: Securix

Originally posted by: jkmccarthy

The real doozie would be if anyone has tried Windows Server 2008 Itanium yet...

If anyone is game to take Securix up on his challenge, apparently you can try for free here:

Download Microsoft Windows Server 2008 for Itanium Trial version

Burning the ISO as I type this...unfortunately I don't think I'll get to trying it til tomorrow...

Well, Windows 2008 Server is being very problematic, at least in my first attempt. I don't have an internal slim DVD drive so I tried my external LG USB DVD drive. It does fire up and does load the Windows installer, though very slowly. It partitions and formats the drive, asks for the CD key (leave it blank for trial) and does the initial unpacking and installation, Then it reboots. So far so good. It gets through all the little green check marks until it gets to the VERY LAST one and gets stuck there forever. I can see the hard disk LED flicker occasionally, but it's been stuck there for at least 20 minutes.

If you get sick of waiting like I did and just reboot the box at that point, it will reboot, get into Windows and say that the Windows install experienced an unexpected error or blah blah and install cannot continue, and it forces you to completely re-install. Running the "repair" function on the install DVD appears to "repair" something but it still dies with that error after rebooting.

I even tried "safe mode" but as soon as the GUI comes up, it gives an error saying that Windows installer cannot finish installing in "safe mode". So much for that.

It's late otherwise I'd let it sit there all night, but the machine is keeping the neighborhood awake.

I also noticed that in the BIOS on booting, sometimes it will see the mouse and report "1 keyboard, 1 mouse" but sometimes just says "1 keyboard". Makes no difference if its cold or warm booted. After Winodws reboots the first time to complete the install, it powers up the mouse (the optical red LED lights up) but the mouse does not respond when moved.

So I'm stuck at that point. I'm gonna scavenge around at work and see if I can liberate an old DVD drive from a retired server somewhere and try that. Other than that, I'm stumped.

One thing's for sure, though. The USB bus on this thing is very flakey. And I could swear it emits microwave radiation because I feel unexplicably warm when it's powered up even though the exhaust is about 4 feet away pointing away from me.

 

[pwfig]

I bought a 80 pin SCSI hard drive plugged in, but the SCSI setup utility cannot detect it, do I need to set SCSI terminator? It seems there is no place to set it. Does anyone use 80 pin SCSI drive too?

I had the same issue until I put SCSI terminators on the two external SCSI connectors on the rear panel of the server, and then it recognized the disk in the scsi select utility.

Also, per the server manual, there is an IDE connector on board, but it seems IDE cable doesn't fit it

The IDE connector which comes to the front of the chassis is for a laptop CD/DVD drive, which I mounted in the chassis and connected to the IDE connector. If you need a full-size IDE connector for a hard drive, you will need to replace the IDE cable provided with a regular hard disk IDE calbe, am\nd make sure the one you get is long enough to reach to the front of the chassis.

 

[Securix]

Also, per the server manual, there is an IDE connector on board, but it seems IDE cable doesn't fit it

The IDE connector which comes to the front of the chassis is for a laptop CD/DVD drive, which I mounted in the chassis and connected to the IDE connector. If you need a full-size IDE connector for a hard drive, you will need to replace the IDE cable provided with a regular hard disk IDE calbe, am\nd make sure the one you get is long enough to reach to the front of the chassis.[/quote]

The unit supports primary and secondary IDE channels. So I'm guessing that the strange very thin IDE-like connector near the cable that connects to the CD/DVD IDE connector is for the other channel. However, I've never seen this type of IDE connector before and am wondering if anyone knows where to find the pinouts for it as well.

Additionally, I have had no further luck installing Windows 2008. One thing I did narrow down is that the hangup that kills the machine after almost the entire config is complete might be the LSI SCSI driver. I downloaded the latest drivers and put them on a USB flash. When prompted in Windows 2008, it DOES accept the drivers. But the minute it loads the drivers, the machine hangs in EXACTLY the same manner as that point towards the end of the install when letting Windows install stock drivers.

 

[pwfig]

During the startup sequence, my Celestica IYA210 frequently hangs after the EFI firmware banner and Broadcom NIC announcement, but before (the second NIC announcement? and) the EFI file systems and boot options menu.

Generally, I found if something wasn't terminated (SCSI connectors) or connected (make sure all 3 LAN interfaces are connected to a switch) or it had trouble recognizing something (my first USB-to-PS/2 dongle) will cause the server to hang or react slowly (not including the delay right after powering on the server).
Has anyone else seen this type of behavior on their IYA210 beast ? Suspected cause and/or workaround ?

Has anyone else tried following the steps in the book to reset the BIOS firmware, and/or update the BIOS firmware ? What pitfalls am I likely to encounter if I try that to cure my machine's tendancy to hang late in the startup sequence ?

I didn't know there was any updated firmware (was it on the CD provided with the server?). Anyway, I didn't try upgrading the BIOS. Let me know how it goes.

Also, has anyone successfully accessed the BMC on their machine via IPMI, and is it possible (just not adviseable?) to reduce the fan speed a notch or two so the machine is not so extremely loud? Or is there a better way to address the noise issue ?

I haven't tried BMC/IPMI yet, but intend to do so. My only concern is the machine runs hot, and historically it has been difficult to cool 1U servers, so I would be concerned about it overheating.
- - -

Lastly, a previous poster mentioned devoting these beasts to BOINC.

I don't know what BOINC is or what it does and haven't tried it. What is it used for?

I've installed the "3rd party" ia64 linux BOINC client, but haven't found any projects with applications that can run on ia64 (itanium) linux.

Good Itanium resources:

Intel.com - CPU info, white papers, links to other Itanium sites,
gelato.org - This is where all the ia64 developer's hang out. White papers, forums, links, etc.
Itanium Solutions Alliance - Sponsored by Intel, it includes all ISVs and hardware vendors which contribute or produce hardware/software based on Itanium (or at least used to),

hp.com - Since HP's Integrity enterprise servers are based on Itanium, they have lots of resources which apply to any Itanium server, not just HP, especially Itanium-specific hardware user manuals.

Redhat.com - The RHEL 4 & 5 Installation Guides have good information about booting Itanium servers. In fact, if you use any Fedora-based Linux, the Redhat documentation generally applies. CentOS links you to the Redhat documentation.

Microsoft.com - Information related to running Windows on IA64, 32-bit compatibility, IA64 Sql Server, lots of technet articles on tuning.

Are you instead using the IA-32 Execution Layer emulation and 32-bit linux BOINC client ? I'd be interested to know how that's working out for you ... my first ia64 linux install on the Celestica was "Scientific Linux IA-64 v4.1" (like CentOS, Sci-Linux is built on Red Hat sources) but v4.1 had the wrong glibc version for the 32-bit linux BOINC client. For this and other reasons, I'm now upgrading to the beta version of Scientific Linux Cern IA-64 v5.x, which includes glibc 2.5.

Anyone running Fedora for Itanium, or is CentOS the ia64 linux distro of choice ?

Since CentOS is a Fedora-based distribution, I figured I had my bases covered, and so far it is largely comaptible with the Redhat releases. I picked CentOS (4) because it had built-in Xen virtualization support, and the goals of the CentOS community is to provide a distro which is stable for server deployment in production environments. The latest Fedora (if there exists an Itanium version) is leading edge, and I didn't want to introduce instability into the environment as Itanium tends to be finicky enough. If they release an Itanium CentOS 5.2 version, I will probably upgrade. I also intend to run Windows Server 2003 with SQL Server (hopefully in a Xen VM) as I need to mess with this environment on occassion.

Other things...

I mentioned in my previous lengthy post that I would upgrade the memory from 2GB (4x512 modules which came with the server) to 8GB (8x1 GB modules). I did the upgrade, and so far, everything is working fine. I bought the memory off eBay for about $400 ($50 per 1 GB module), making sure it met the specs mentioned in the user manual (ECC Registered, DDR SDRAM, etc).

Lastly, I'm trying to figure out how to re-route and wire an on/off switch to the front of the chassis as it is really inconvenient to get behind the chassis since it sits in a rack. I though I would be clever and connect the power to a web-based power switch where I could toggle the power remotely to enable the server, but it works inconsistently. Wake-On-LAN would be another possibility I haven't explored. Anybody got any ideas, including the proper type and specs for a repalcement switch?

 

[pwfig]

I have a SIIG "active" PS2 to USB adaptor that connects PS2 keyboard and PS2 mouse into one of the four USB inputs in the back of the server...

The USB-PS/2 dongle which seems to work consistently for me is called the "Y-Mouse" which is manufactured by PI Engineering. I later switched to a generic dongle (nine dollar special at Fry's) which caused the hanging, so I went back to the Y-Mouse and left it there.

As far as keyboard and mouse, I use a Logitech Wireless keyboard and Logitech Cordless Trackman Wheel connected to a 30 foot KVM cable going into a Compaq 8-port KVM, and haven't had any wierdness. In fact, the mouse behaves better on the Itanium over my x86 CentOS 5 servers where I get jumpiness on the cursor, but may have figured that out when I apply some recommended driver settings at the Xorg site.

 

[pwfig]

Well, Windows 2008 Server is being very problematic...

I haven't gotten around yet to trying Windows, but I may push this up since you are giving it a shot. I will try Windows Server 2003 first to eliminate the possibility that the 2008 version may be flakey due to its "newness" and may not have all the drivers you think it should have (anybody here try to load Vista without problems?).

I've worked on some other Itanium servers where a driver disk was required during the initial install of Windows, and I think using an IDE-based CD drive will help get you away from any potential USB flakiness.

Where I really intend to go with this (a glutton for punishment?) is to make the Itanium server boot from an iSCSI LUN (one each for Centos and Windows) to a NAS server I built using the OpenFiler software, then make LUNs for Xen-based Itanium VMs.

 

[pwfig]

Here's the details on my machine ...

American Megatrends Enterprise64
BIOS 12/08/2003 10:06:38

... (then successfully completes POST) ...

LSI Logic Corp, MPT BIOS

MPT-BIOS 5.07.00
... (then successfully scans SCSI bus and identifies drives) ...

EFI version 1.10 [14.61]
EFI IA-64 SDV/FDK (BIOS Callbacks) [Mon Dec 8 10:07:07 2003]
This image MainEntry is at 000000007FA02000

Copyright (c) 2000-2003 Broadcom Corporation
Broadcom NetXtreme Gigabit Ethernet EFI driver v6.1.2


_ <== more often than not, hangs with cursor here

All my revisions are identical to yours (BIOS, LSI BIOS, EFI, Broadcom EFI driver, etc).
I was going to time the delay when initializing the NICs, but it flashed by too quickly (no hang or delay) until it got to the EFI pause countdown, then advacnes to the EFI boot screen with Centos as the first choice, and loads Centos after a few seconds delay.

Other than the USB issue previously mentioned, I've got all three Ethernet ports plugged into a Linksys SRW2024 gigabit switch, where the two gig ports are connecting at gigabit speed, and the management port - I don't know the speed setting.

After CentOS starts loading, I get a non-fatal error/warning:
"get ACPI_serial_port zero_length_to IO Port Range?"

I tried messing with the BIOS serial interface settings, which made no difference. Since it mentions ACPI, disabling APIC during boot (add NOAPIC to commandline boot options in elilo) may fix this.

After doing a dmesg from within Centos, there are a few strange things which probably need addressing to get a completely clean boot. I included the dmesg output here as some of the info shown may be helpful (errors and warnings highlighted in italics):

[root@localhost ~]# dmesg
Linux version 2.6.9-67.0.22.EL (builder@c4ia64) (gcc version 3.4.6 20060404 (Red Hat 3.4.6-9)) #1 SMP Sat Jul 26 19:11:08 EEST 2008
EFI v1.10 by INTEL: SALsystab=0x7fe4d030 ACPI=0x7ff9a000 ACPI 2.0=0x7ff99000 SMBIOS=0xf00a0
booting generic kernel on platform dig
ACPI: RSDP (v002 AMI ) @ 0x000000007ff99000
ACPI: XSDT (v001 AMI CA_MOUNT 0x01072002 MSFT 0x00010013) @ 0x000000007ff99090
ACPI: FADT (v003 AMI CA_MOUNT 0x01072002 MSFT 0x00010013) @ 0x000000007ff99138
ACPI: MADT (v001 AMI CA_MOUNT 0x01072002 MSFT 0x00010013) @ 0x000000007ff99230
ACPI: IPPT (v001 AMI CA_MOUNT 0x01072002 MSFT 0x00010013) @ 0x000000007ff99328
ACPI: SPCR (v001 AMI CA_MOUNT 0x01072002 MSFT 0x00010013) @ 0x000000007ff99360
ACPI: DSDT (v001 AMI CA_MOUNT 0x00000000 MSFT 0x0100000e) @ 0x0000000000000000
Warning: acpi_table_parse(ACPI_SRAT) returned 0!
Warning: acpi_table_parse(ACPI_SLIT) returned 0!
efi.trim_top: ignoring 4KB of memory at 0x0 due to granule hole at 0x0
efi.trim_top: ignoring 488KB of memory at 0x1000 due to granule hole at 0x0
efi.trim_top: ignoring 148KB of memory at 0x7b000 due to granule hole at 0x0
efi.trim_bottom: ignoring 15360KB of memory at 0x100000 due to granule hole at 0x0
Initial ramdisk at: 0xe0000002fedcd000 (1896148 bytes)
SAL 3.1: AMERICAN MEGATRENDS INC, (C)AMI, 64-BIT AMI-BIOS(C)AMI 1985-2003 version 3.0
SAL Platform features: BusLock IRQ_Redirection
SAL: AP wakeup using external interrupt vector 0xf0
No logical to physical processor mapping available
iosapic_system_init: Disabling PC-AT compatible 8259 interrupts
ACPI: Local APIC address c0000000fee00000
PLATFORM int CPEI (0x3): GSI 22 (level, low) -> CPU 0 (0xf81f) vector 30
register_intr: changing vector 39 from IO-SAPIC-edge to IO-SAPIC-level
2 CPUs available, 2 CPUs total
Registering legacy COM ports for serial console
MCA related initialization done
Virtual mem_map starts at 0xa0007ffffd600000
On node 0 totalpages: 522088
DMA zone: 129042 pages, LIFO batch:4
Normal zone: 393046 pages, LIFO batch:4
HighMem zone: 0 pages, LIFO batch:1
Built 1 zonelists
Kernel command line: BOOT_IMAGE=scsi0:EFI\redhat\vmlinuz-2.6.9-67.0.22.EL rhgb quiet root=LABEL=/ ro
PID hash table entries: 4096 (order: 12, 131072 bytes)
CPU 0: base freq=200.022MHz, ITC ratio=14/2, ITC freq=1400.158MHz+/--1ppm
Console: colour VGA+ 80x25
Dentry cache hash table entries: 1048576 (order: 9, 8388608 bytes)
Inode-cache hash table entries: 524288 (order: 8, 4194304 bytes)
Placing software IO TLB between 0x4aa8000 - 0x8aa8000
Memory: 8240704k/8353408k available (5781k code, 124800k reserved, 2346k data, 384k init)
McKinley Errata 9 workaround not needed; disabling it
Calibrating delay loop... 2098.00 BogoMIPS (lpj=1024000)
Security Scaffold v1.0.0 initialized
SELinux: Initializing.
SELinux: Starting in permissive mode
There is already a security framework initialized, register_security failed.selinux_register_security: Registering secondary module capability
Capability LSM initialized as secondary
Mount-cache hash table entries: 1024 (order: 0, 16384 bytes)
Boot processor id 0x0/0xf81f
task migration cache decay timeout: 10 msecs.
CPU 1: synchronized ITC with CPU 0 (last diff 0 cycles, maxerr 521 cycles)
CPU 1: base freq=200.022MHz, ITC ratio=14/2, ITC freq=1400.158MHz+/--1ppm
Calibrating delay loop... 2094.88 BogoMIPS (lpj=1021952)
Brought up 2 CPUs
Total of 2 processors activated (4192.88 BogoMIPS).
checking if image is initramfs... it is
Freeing initrd memory: 1824kB freed
NET: Registered protocol family 16
ACPI: Subsystem revision 20040816
ACPI: Interpreter enabled
ACPI: Using IOSAPIC for interrupt routing
ACPI: PCI Root Bridge [PCI0] (00:00)
PCI: Ignoring BAR0-3 of IDE controller 0000:00:1f.1
ACPI: PCI Interrupt Routing Table [\_SB_.PCI0._PRT]
ACPI: PCI Interrupt Routing Table [\_SB_.PCI0.H2PB._PRT]
ACPI: PCI Root Bridge [PCI1] (00:02)
ACPI: PCI Interrupt Routing Table [\_SB_.PCI1.P2PA._PRT]
ACPI: PCI Interrupt Routing Table [\_SB_.PCI1.P2PB._PRT]
ACPI: PCI Root Bridge [PCI2] (00:05)
ACPI: PCI Interrupt Routing Table [\_SB_.PCI2.P2PA._PRT]
ACPI: PCI Interrupt Routing Table [\_SB_.PCI2.P2PB._PRT]
ACPI: Device [CSFF] status [00000008]: functional but not present; setting present
ACPI: PCI Root Bridge [CSFF] (00:ff)
usbcore: registered new driver usbfs
usbcore: registered new driver hub
PCI: Using ACPI for IRQ routing
GSI 16 (level, low) -> CPU 0 (0xf81f) vector 48
ACPI: PCI Interrupt 0000:00:1d.0[A] -> GSI 16 (level, low) -> IRQ 48
GSI 19 (level, low) -> CPU 0 (0xf81f) vector 49
ACPI: PCI Interrupt 0000:00:1d.1 -> GSI 19 (level, low) -> IRQ 49
GSI 18 (level, low) -> CPU 0 (0xf81f) vector 50
ACPI: PCI Interrupt 0000:00:1d.2[C] -> GSI 18 (level, low) -> IRQ 50
GSI 23 (level, low) -> CPU 0 (0xf81f) vector 51
ACPI: PCI Interrupt 0000:00:1d.7[D] -> GSI 23 (level, low) -> IRQ 51
ACPI: PCI interrupt 0000:00:1f.1[A]: no GSI
ACPI: PCI interrupt 0000:00:1f.3: no GSI
ACPI: PCI Interrupt 0000:01:09.0[A] -> GSI 16 (level, low) -> IRQ 48
GSI 48 (level, low) -> CPU 0 (0xf81f) vector 52
ACPI: PCI Interrupt 0000:03:01.0[A] -> GSI 48 (level, low) -> IRQ 52
GSI 96 (level, low) -> CPU 0 (0xf81f) vector 53
ACPI: PCI Interrupt 0000:06:01.0[A] -> GSI 96 (level, low) -> IRQ 53
GSI 100 (level, low) -> CPU 0 (0xf81f) vector 54
ACPI: PCI Interrupt 0000:06:02.0[A] -> GSI 100 (level, low) -> IRQ 54
GSI 72 (level, low) -> CPU 0 (0xf81f) vector 55
ACPI: PCI Interrupt 0000:07:01.0[A] -> GSI 72 (level, low) -> IRQ 55
GSI 73 (level, low) -> CPU 0 (0xf81f) vector 56
ACPI: PCI Interrupt 0000:07:01.1 -> GSI 73 (level, low) -> IRQ 56
perfmon: version 2.0 IRQ 238
perfmon: Itanium 2 PMU detected, 16 PMCs, 18 PMDs, 4 counters (47 bits)
PAL Information Facility v0.5
perfmon: added sampling format default_format
perfmon_default_smpl: default_format v2.0 registered
audit: initializing netlink socket (disabled)
audit(1223564077.452:1): initialized
Total HugeTLB memory allocated, 0
VFS: Disk quotas dquot_6.5.1
Dquot-cache hash table entries: 2048 (order 0, 16384 bytes)
SELinux: Registering netfilter hooks
Initializing Cryptographic API
ksign: Installing public key data
Loading keyring
- Added public key 7E9897C58FA53CD0
- User ID: CentOS (Kernel Module GPG key)
pci_hotplug: PCI Hot Plug PCI Core version: 0.5
ACPI: Processor [CPU0] (supports C1)
ACPI: Processor [CPU1] (supports C1)
EFI Time Services Driver v0.4
Linux agpgart interface v0.100 (c) Dave Jones
serio: i8042 AUX port at 0x60,0x64 irq 36
serio: i8042 KBD port at 0x60,0x64 irq 32
Serial: 8250/16550 driver $Revision: 1.90 $ 64 ports, IRQ sharing enabled
acpi_serial_port: zero-length IO port range?ttyS0 at I/O 0x3f8 (irq = 45) is a 16550A
RAMDISK driver initialized: 16 RAM disks of 16384K size 1024 blocksize
divert: not allocating divert_blk for non-ethernet device loUniform Multi-Platform E-IDE driver Revision: 7.00alpha2
ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xxICH4: IDE controller at PCI slot 0000:00:1f.1
PCI: Device 0000:00:1f.1 not available because of resource collisions
ACPI: PCI interrupt 0000:00:1f.1[A]: no GSI
ICH4: BIOS configuration fixed.
ICH4: chipset revision 2
ICH4: not 100% native mode: will probe irqs later
ide0: BM-DMA at 0x1000-0x1007, BIOS settings: hdaio, hdbio
ide1: BM-DMA at 0x1008-0x100f, BIOS settings: hdcio, hddMA
Probing IDE interface ide0...
Probing IDE interface ide1...
hdd: TEAC DVD-ROM DV-28E-C, ATAPI CD/DVD-ROM drive
Using cfq io scheduler
ide1 at 0x170-0x177,0x376 on irq 33
Probing IDE interface ide0...
Probing IDE interface ide2...
Probing IDE interface ide3...
Probing IDE interface ide4...
Probing IDE interface ide5...
hdd: ATAPI 24X DVD-ROM drive, 256kB Cache, UDMA(33)
Uniform CD-ROM driver Revision: 3.20
ide-floppy driver 0.99.newide
usbcore: registered new driver hiddev
usbcore: registered new driver usbhid
drivers/usb/input/hid-core.c: v2.0:USB HID core driver
mice: PS/2 mouse device common for all mice
atkbd.c: keyboard reset failed on isa0060/serio1
atkbd.c: keyboard reset failed on isa0060/serio0
md: md driver 0.90.0 MAX_MD_DEVS=256, MD_SB_DISKS=27
EFI Variables Facility v0.08 2004-May-17
NET: Registered protocol family 2
IP route cache hash table entries: 262144 (order: 7, 2097152 bytes)
TCP established hash table entries: 1048576 (order: 10, 16777216 bytes)
TCP bind hash table entries: 1048576 (order: 10, 16777216 bytes)
TCP: Hash tables configured (established 1048576 bind 1048576)
Initializing IPsec netlink socket
NET: Registered protocol family 1
NET: Registered protocol family 17
Freeing unused kernel memory: 384kB freed
SCSI subsystem initialized
Fusion MPT base driver 3.02.99.00rh
Copyright (c) 1999-2007 LSI Logic Corporation
Fusion MPT SPI Host driver 3.02.99.00rh
ACPI: PCI Interrupt 0000:07:01.0[A] -> GSI 72 (level, low) -> IRQ 55
mptbase: Initiating ioc0 bringup
ioc0: 53C1030: Capabilities={Initiator}
scsi0 : ioc0: LSI53C1030, FwRev=01030100h, Ports=1, MaxQ=255, IRQ=55
ACPI: PCI Interrupt 0000:07:01.1 -> GSI 73 (level, low) -> IRQ 56
mptbase: Initiating ioc1 bringup
ioc1: 53C1030: Capabilities={Initiator}
scsi1 : ioc1: LSI53C1030, FwRev=01030100h, Ports=1, MaxQ=255, IRQ=56
Vendor: HP Model: 36.4GB C 80-F612 Rev:
Type: Direct-Access ANSI SCSI revision: 02
SCSI device sda: 71132960 512-byte hdwr sectors (36420 MB)
SCSI device sda: drive cache: write back
SCSI device sda: 71132960 512-byte hdwr sectors (36420 MB)
SCSI device sda: drive cache: write back
sda: sda1 sda2 sda3
Attached scsi disk sda at scsi1, channel 0, id 3, lun 0
Fusion MPT SAS Host driver 3.02.99.00rh
libata version 2.00 loaded.
kjournald starting. Commit interval 5 seconds
EXT3-fs: mounted filesystem with ordered data mode.
SELinux: Disabled at runtime.
SELinux: Unregistering netfilter hooks
tg3.c:v3.77 (May 31, 2007)
ACPI: PCI Interrupt 0000:06:01.0[A] -> GSI 96 (level, low) -> IRQ 53
divert: allocating divert_blk for eth0
eth0: Tigon3 [partno(BCM95703A30) rev 1002 PHY(5703)] (PCIX:133MHz:64-bit) 10/100/1000Base-T Ethernet 00:e0:ec:13:3b:20
eth0: RXcsums[1] LinkChgREG[0] MIirq[0] ASF[0] WireSpeed[1] TSOcap[1]
eth0: dma_rwctrl[769c4000] dma_mask[64-bit]
ACPI: PCI Interrupt 0000:06:02.0[A] -> GSI 100 (level, low) -> IRQ 54
divert: allocating divert_blk for eth1
eth1: Tigon3 [partno(BCM95703A30) rev 1002 PHY(5703)] (PCIX:133MHz:64-bit) 10/100/1000Base-T Ethernet 00:e0:ec:13:3b:22
eth1: RXcsums[1] LinkChgREG[0] MIirq[0] ASF[0] WireSpeed[1] TSOcap[1]
eth1: dma_rwctrl[769c4000] dma_mask[64-bit]
e100: Intel(R) PRO/100 Network Driver, 3.5.10-k2-NAPI
e100: Copyright(c) 1999-2005 Intel Corporation
ACPI: PCI Interrupt 0000:03:01.0[A] -> GSI 48 (level, low) -> IRQ 52
divert: allocating divert_blk for eth2
e100: eth2: e100_probe: addr 0xec0ff000, irq 52, MAC addr 00:E0:EC:13:3B:7A
ACPI: PCI Interrupt 0000:00:1d.7[D] -> GSI 23 (level, low) -> IRQ 51
ehci_hcd 0000:00:1d.7: EHCI Host Controller
ehci_hcd 0000:00:1d.7: irq 51, pci mem c0000000fdfffc00
ehci_hcd 0000:00:1d.7: new USB bus registered, assigned bus number 1
PCI: slot 0000:00:1d.7 has incorrect PCI cache line size of 0 bytes, correcting to 128
ehci_hcd 0000:00:1d.7: USB 2.0 enabled, EHCI 1.00, driver 2004-May-10
hub 1-0:1.0: USB hub found
hub 1-0:1.0: 6 ports detected
USB Universal Host Controller Interface driver v2.2
ACPI: PCI Interrupt 0000:00:1d.0[A] -> GSI 16 (level, low) -> IRQ 48
uhci_hcd 0000:00:1d.0: UHCI Host Controller
uhci_hcd 0000:00:1d.0: irq 48, io base 000000000000df20
uhci_hcd 0000:00:1d.0: new USB bus registered, assigned bus number 2
hub 2-0:1.0: USB hub found
hub 2-0:1.0: 2 ports detected
ACPI: PCI Interrupt 0000:00:1d.1 -> GSI 19 (level, low) -> IRQ 49
uhci_hcd 0000:00:1d.1: UHCI Host Controller
uhci_hcd 0000:00:1d.1: irq 49, io base 000000000000df40
uhci_hcd 0000:00:1d.1: new USB bus registered, assigned bus number 3
hub 3-0:1.0: USB hub found
hub 3-0:1.0: 2 ports detected
ACPI: PCI Interrupt 0000:00:1d.2[C] -> GSI 18 (level, low) -> IRQ 50
uhci_hcd 0000:00:1d.2: UHCI Host Controller
uhci_hcd 0000:00:1d.2: irq 50, io base 000000000000df80
uhci_hcd 0000:00:1d.2: new USB bus registered, assigned bus number 4
hub 4-0:1.0: USB hub found
hub 4-0:1.0: 2 ports detected
usb 3-2: new low speed USB device using address 2
md: Autodetecting RAID arrays.
md: autorun ...
md: ... autorun DONE.
input: USB HID v1.00 Keyboard [P.I. Engineering PC Keyboard/Mouse to USB Adapter] on usb-0000:00:1d.1-2
input: USB HID v1.00 Mouse [P.I. Engineering PC Keyboard/Mouse to USB Adapter] on usb-0000:00:1d.1-2
NET: Registered protocol family 10
Disabled Privacy Extensions on device a00000010078f518(lo)
IPv6 over IPv4 tunneling driver
divert: not allocating divert_blk for non-ethernet device sit0
ACPI: Power Button (FF) [PWRF]
EXT3 FS on sda3, internal journal
device-mapper: 4.5.5-ioctl (2006-12-01) initialised: dm-devel@redhat.com
device-mapper: dm-multipath version 1.0.5 loaded
Adding 1534160k swap on /dev/sda2. Priority:-1 extents:1
ip_tables: (C) 2000-2002 Netfilter core team
ip_conntrack version 2.1 (8192 buckets, 65536 max) - 456 bytes per conntrack
Linux Kernel Card Services
options: [pci] [cardbus] [pm]
ADDRCONF(NETDEV_UP): eth0: link is not ready
tg3: eth0: Link is up at 1000 Mbps, full duplex.
tg3: eth0: Flow control is off for TX and off for RX.
ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready
ADDRCONF(NETDEV_UP): eth1: link is not ready
tg3: eth1: Link is up at 1000 Mbps, full duplex.
tg3: eth1: Flow control is off for TX and off for RX.
ADDRCONF(NETDEV_CHANGE): eth1: link becomes ready
ADDRCONF(NETDEV_UP): eth2: link is not ready
e100: eth2: e100_watchdog: link up, 100Mbps, full-duplex
ADDRCONF(NETDEV_CHANGE): eth2: link becomes ready
eth0: no IPv6 routers present
eth1: no IPv6 routers present
eth2: no IPv6 routers present
lp: driver loaded but no devices found
[root@localhost ~]#

 

[Securix]

I updated all the flashable items that were provided on the CD. Unfortunately what is on the CD is the same as what is already on the box and what is (looking at everyone's BIOS revs) apparently on everyone's box.

I haven't tried Windows 2003. Maybe I'll give it a shot over the weekend. Possibly if it works, upgrading to 2008 might work as well.

 

[jkmccarthy]

Thanks pwfig for the informative replies ... I don't know what became of the e-mail notices I used to receive each time a new post appears in this forum ... sorry for the delay in saying thanks and replying ...

The USB-PS/2 dongle which seems to work consistently for me is called the "Y-Mouse" which is manufactured by PI Engineering

The USB "Y-mouse" is rather pricey, so I'm still keeping an eye out for a plain generic USB keyboard. Meanwhile, I've discovered that running my existing keyboard through a Belkin SOHO (small office / home office) USB KVM seems to make the Celestica happy about the keyboard+mouse, so it's been booting just fine now for the past month.

In case you're still curious, BOINC = Berkeley Open Infrastructure for Network Computing ... it's a software package that allows your computer to take part in highly parallel distributed computing projects like SETI at Home, Einstein at Home, ClimatePrediction Dot Net (which is my interest lately), etc. But as stated not many (not any?) of these projects make their BOINC client application programs available for ia64, even though the BOINC server program apparently does run on itanium.

I'm currently running a RHEL 5 -based Linux, but may need to switch to Fedora-9 or Fedora-10-beta for ia64 before getting the righ combination of BOINC software + the IA-32 Execution Layer emulation (to let me run x86 BOINC application clients) to get this to work. I've tried a few times to do a net install of the Fedora-10-beta (the so-called "rawhide" version...), but after a few failed installs (and learning from my mistakes, getting a little further each time), the last error message I encountered said something about it not being able to find the kernel-smp rpm, so at that point I decided "rawhide" was probably a little to raw for me....

So how is everyone else getting along now with their Celestica IYA210 beasts ??

-- Jim

 

[Zan Lynx]

I finally got off my butt and installed Fedora 9 on mine. I would have done 10 or 11-alpha but no one seems to be building Fedora for IA64 any more. Sad. Perhaps I'll fix that.

I was going to install Gentoo but their IA64 installer CD wouldn't boot the kernel after loading initrd.gz

I seem to be having the same problem as described above with the SCSI drive going into narrow mode. I'll have to find some terminators for the back. I guess the card can't be set for internal termination of that port? I haven't looked honestly. It takes sooo long to boot. Probably the ECC setup. And if I tweak anything it just forgets it again since the CMOS battery is dead.

I've got a frackin' Tyan dual socket Pentium 166 from 1996 still running and IT never complains about a low CMOS battery.

Lot's of work left for me to do on this I see.

 

[Nothinman]

but no one seems to be building Fedora for IA64 any more. Sad. Perhaps I'll fix that.

Good luck, there's a reason that most people have abandoned IA64...

 

[jhu]

Originally posted by: Nothinman
Good luck, there's a reason that most people have abandoned IA64...

i'd say it's an issue of low availability. if you're going to use some sort of linux-based os, what's the point in using esoteric hardware when any cheapo x86 white box computer will do?

i used to have several macs and sparcstations and loaded netbsd on those things just to mess around. in the end i got rid of them because they were just wasting electricity.

otoh, people would come over and wonder what the hell all these computers were doing in my room.

 

[Zan Lynx]

I agree about availability. It seems nearly impossible and far too expensive to just go to HP.com and order up an Itanium workstation.

I believe that if more Linux users tried them, they would like them.

One of the benefits that I have seen is very smooth multitasking. Admittedly, a Xeon with SCSI drives is also very smooth. I suppose I have to attribute it to either huge CPU cache sizes or SCSI.

 

[Nothinman]

I believe that if more Linux users tried them, they would like them.

Why? Out of all of the Linux servers at my last job the handful of IA64 boxes we had were the most problematic.

 

[Zan Lynx]

First, I don't know what you mean by "problematic." From my experience with one other IA64, they aren't. I have a Compaq DL/590 and it has never given me trouble, once I got it set up (it's heavy!) and got past the bugs of the Debian IA64 installer. Those bugs have since been fixed.

Second, here is why I think Linux users would like them: I am a Linux user and I like them. Two friends of mine who also do Linux programming look at IA64 and go, "Oooh, cool, I like that." although they haven't bought one.

I will expand on this a bit:

I enjoy using new and different equipment. So do a lot of Linux users. That is one big reason they are using Linux or BSD in the first place instead of Windows Vista Home Extra Special.

Also, I believe Itanium is better in a lot of ways than Xeon. There's less CPU dedicated to fixing up all the legacy mistakes of x86 and translating it to the real CPU micro-op instructions. That leaves more room for registers and cache.

Admittedly, x86 crap doesn't show much on the surface because Intel has done a really good job of shoveling it under the covers and so a Xeon is a perfectly good CPU.

If you only care about the end results, I suppose there's not much reason to use IA64 for most applications.

For myself, I would like the x86 architecture to shrivel up and die eventually and I think IA64 is a better future design than x86_64, even though most of my systems are now running x86_64 (two laptops, two desktops, all 64-bit Windows or Linux).

 

[Nothinman]

First, I don't know what you mean by "problematic."

Constant but varied problems, 3 in particular were always having issues. They were an Oracle cluster with 1 being a backup node at another site. IIRC they were HP RX2600s. We had HP there swapping out hardware left and right. All of the x86 HP and Dell boxes we had on that network combined probably had less problems.

Second, here is why I think Linux users would like them: I am a Linux user and I like them.

I like EFI better than legacy BIOS but that's not enough reason to choose them IMO. Although the fact that EFI requires a FAT partition kinda sucks.

Two friends of mine who also do Linux programming look at IA64 and go, "Oooh, cool, I like that." although they haven't bought one.

And then after you actually use it you realize that the architecture really isn't that important. I had Debian on 4 different architectures at my house and it worked exactly the same on all of them. Which is the whole point, once you get past the "Wow, SRM/EFI/OBP is really nice compared to BIOS" and you get the thing installed there's not much more to think about because it's just another Linux install.

Also, I believe Itanium is better in a lot of ways than Xeon. There's less CPU dedicated to fixing up all the legacy mistakes of x86 and translating it to the real CPU micro-op instructions. That leaves more room for registers and cache.

Too bad that the IA64 instruction set is so complicated that it takes an Intel engineer to write assembly for it so if the compiler doesn't do the right thing it's noticably slower than x86.

If you only care about the end results, I suppose there's not much reason to use IA64 for most applications.

Which is kind of the point. It's fun to play with if you've never seen one, but the novelty wears off fairly quickly.

For myself, I would like the x86 architecture to shrivel up and die eventually and I think IA64 is a better future design than x86_64, even though most of my systems are now running x86_64 (two laptops, two desktops, all 64-bit Windows or Linux).

IMO IA64 is essentially dead outside of some very niche places that need the high availability options that aren't there yet in AMD64 hardware and that won't last forever.