Linux Equivalent to HP-UX commands

[Vinesh Dhevcharran]

Hi,

New Linux user on board.

1. On HP-UX I use lvdisplay to check the status of the raid devices. What can I use on Linux is the disks are raided with a raid controller?

2. To find the number of active CPUs I use
ioscan -fCprocessor. What can I use on Linux, except top?

3. I have an HP Ultrium 230 on an IBX x225 on Redhat 7.3. After a few backups, the tape drive is no longer available on the OS. After a reboot, the tape drive is back. Any ideas as to what the problem here is?

Many Thanks

[Balaji N]

1. no idea.
2. cat /proc/cpuinfo.
3. is there some kind of loose connection. any errors in dmesg / var/log/messages?

hth
-balaji

[Ramkumar Devanathan]

hi,

1. you can still use lvdisplay if you've configured lvm and volumes on the redhat 7.3 machine.

2. cat /proc/cpuinfo

3. any messages in /var/log/messages or any other logfile?

- ramd.

[John Meissner]

I found this on the internet and hopefully it's not too big to attach. It's called the rosetta stone and it has many cross platform references to commands.

[Bill Douglass]

If you have a h/w RAID controller, there is likely a BIOS interface you can use to view/modify the volumes under its control. For Adaptec/PERC RAID controllers, you would type a Ctrl-A when prompted during boot-up. What type of RAID controller do you have?

As mentioned in other messages, /proc/cpuinfo gives you lots of detail about the CPU configuration fo your system.

Keep in mind that Intel Xeon chips with the hyper-threading capability will show up by default as two CPUs per physical chip; this is intentional as it allows the system to take advantage of the parallelism built into the hyper-threaded chips.

Regardign the tape drive, what back-up s/w are you using? Are there any error messages logged in /var/adm/messages?

[dirk dierickx]

1. lvdisplay will still work, if you are using LVM ofcourse. the status of a software raid setup can be viewed with lsraid. use fdisk for your basic disk partition information

2. cat /proc/cpu

3. perhaps the driver for your tapedrive is a module that gets unloaded after some time. check with 'lsmod' if the tape drive driver is still loaded, if not do a 'modprobe' to load it again, check if it works after this. if the driver is loaded and the drive won't work, it might be a good idea to unload it with 'rmmod' and load it again after that with 'modprobe'.

[Jarle Bjorgeengen]

Hi,

1. I think it sounds strange to use lvdisplay on hp-ux to check the status of raid devices.

Normally I would use lvdisplay -v in hpux to see if there are stale extents when using mirrordisk UX. Mirrordisk UX is software raid 1.

In Linux there is no software raid buildt into LVM. You have to use MultiDisk devivces to get software raid. You can get both raid 0, 1 and 5 with md. To also use LVM with your software raid devices , you use the md devices to make up your LVM volume group.

Now, if you have a raid controller, you dont need to worry about these things. Alle is taken care of by the controller. You need of course to define logical devices to presnt to the OS, but that is done in a separate controller BIOS.(Or special driver software) If the device driver of the raid controller works properly, You will get /dev/sda--x devices for each logical device you have, and you can use them directly to put a filesystem on, or you can put them into a LVM volumegroup. Anyway using a raid controller , the data redundancy is thaken care of by the controller hw.

Oh BTW , you can use hw-raidcontrollers in hp-ux also, and you dont need Mirrordisk UX.

A virtual array, typically has a raid controller inside , and presents logical devices to the OS.

Rgds Jarle

[david_69]

What sort of HBA do you have?

try:
lspci
dmesg
more /proc/ioports
ls -la /proc/scsi

The way you manipulate Hardware raid volumes is vendor specific.

Here is one example of the process for playing with raid volumes:
http://www.dandelion.com/Linux/README.DAC960

[Vinesh Dhevcharran]

Hi,

Many thanks for the responses.

The tape errors that I found in the messages files are below:

st0: Error 70000 (sugg. bt 0x0, driver bt 0x0, host
bt 0x7).
Apr 16 15:45:06 SMUG kernel: st0: Error with sense data: Current st09:00: sense
key Unit Attention
Apr 16 15:45:06 SMUG kernel: Additional sense indicates Power on,reset,or bus de
vice reset occurred
Apr 16 15:45:06 SMUG kernel: st0: Error on write filemark.

Also, what I noticed is that the cpio is taking approx. 5GB per hour. For an LTO this is very slow. My backup statement is as follows:
find . -print | cpio -ocvB -F /dev/st0

Any suggestions on how I can speed this up?

[david_69]

On you performance issue:
What is you disk performance like?
Try vmstat, also take a look at /proc/stat ( There should be a line called disk_io that can give you per disk stats).
You might also be able to get good "per disk" stats from a file in /proc/scsi ( Again this is vendor specific) YMMV.
What is your CPU upto as you do the cpio ? What does the vendor say your block size should be? Check out www.linuxtapecert.org.

Hope this helps.