j-g-faustus's questions

I'm getting "NVRM: os_raise_smp_barrier()" messages in syslog and dmesg at a rate of 4-6 per minute; I think it's from the Nvidia graphics driver.

The system behaves normally, but the messages clutter up the logs so I can't find any real issues. (At the moment dmesg is 4,000 lines and 3,925 of those lines are the same message.)

Is it possible to suppress or redirect these messages?

Example log snippet:

Mar 13 20:38:04 desktop kernel: [175481.413559] NVRM: os_raise_smp_barrier(), invalid context!
Mar 13 20:38:04 desktop kernel: [175481.430904] NVRM: os_raise_smp_barrier(), invalid context!
Mar 13 20:38:45 desktop kernel: [175522.223337] NVRM: os_raise_smp_barrier(), invalid context!
Mar 13 20:38:45 desktop kernel: [175522.239655] NVRM: os_raise_smp_barrier(), invalid context!
Mar 13 20:38:47 desktop kernel: [175523.556706] NVRM: os_raise_smp_barrier(), invalid context!
Mar 13 20:38:47 desktop kernel: [175523.572015] NVRM: os_raise_smp_barrier(), invalid context!

Nautilus seems to list all partitions under "devices".

On my system, this includes a Windows 7 partition and a "System Reserved" partition created by Windows.

I would prefer to hide those two partitions. Is it possible?

Related to a question on which is better, the CFS or BFS scheduler:

Where do I find a list of which schedulers are available?

And how do I switch between them?

I have a few hard drives in mdadm RAID 5 configured to go to standby after a few minutes of inactivity. (Using hdparm.conf spindown_time.)

At irregular intervals I get messages like these in dmesg:

[ 1840.251661] ata4.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen
[ 1840.251722] ata4.00: failed command: SMART
[ 1840.251758] ata4.00: cmd b0/d5:01:06:4f:c2/00:00:00:00:00/00 tag 0 pio 512 in
[ 1840.251759]          res 40/00:14:50:2e:04/00:00:02:00:00/40 Emask 0x4 (timeout)
[ 1840.251858] ata4.00: status: { DRDY }
[ 1840.251888] ata4: hard resetting link
[ 1840.600742] ata4: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
[ 1840.601521] ata4.00: configured for UDMA/133
[ 1840.601547] ata4: EH complete
[337877.713988] ata4.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen
[337877.714019] ata4.00: failed command: SMART
[337877.714038] ata4.00: cmd b0/d5:01:06:4f:c2/00:00:00:00:00/00 tag 0 pio 512 in
[337877.714039]          res 40/00:04:90:10:81/00:00:00:00:00/40 Emask 0x4 (timeout)
[337877.714089] ata4.00: status: { DRDY }
[337877.714107] ata4: hard resetting link
[337878.063085] ata4: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
[337878.063743] ata4.00: configured for UDMA/133
[337878.063764] ata4: EH complete

I think the exception is caused by smartd when a drive does not wake up quickly enough.

There are no issues (that I can tell) in accessing the drives normally through the file system - it takes a few seconds longer than normal when they are asleep, but there are no exceptions.

Is this something I should worry about, as a potential symptom on something that could corrupt a drive over time?

Or can I safely ignore it as part of normal operation?

Edit:

By request: smartctl -a for sdaand sde, both disks are members of the array.

If ata4is the same as scsi-4 then sde is the one that gave the error above, according to /dev/disk/by-path.

Using findsmb on my home network gives me this:

IP ADDR         NETBIOS NAME     WORKGROUP/OS/VERSION 
---------------------------------------------------------------------
192.168.1.3     MINI           [MSHOME] [Unix] [Samba 3.0.25b-apple]
192.168.1.9     SERVER        +[    MSHOME        ]
192.168.1.11    NAS            [    WORKGROUP     ]

• MINI - a Mac mini
• SERVER - Ubuntu 10.10 server; Samba installed but not running.
• NAS - another Ubuntu 10.10 server; Samba installed but not running.

When Samba is running the Ubuntu groups are the values from /etc/samba/smb.conf, but with Samba turned off it falls back to these defaults - even different defaults for two installations of the same OS.

I would like to change the workgroups, but I can't figure out where the value is coming from: Not from Samba (that's a different value), not from Winbind (not running), not from Avahi (not installed); and I don't know of anything else involved in network groups.

How can I find this setting so I can change it?

How do I best update lm-sensors to work with recent hardware?

I have a fresh install of Ubuntu 10.10 server, and get the following:

$ sudo apt-get install lm-sensors
$ sensors
No sensors found! [...] 
Try sensors-detect [...]
$ sudo sensors-detect
[...]
Driver 'to-be-written':
  * Chip "Nuvoton W83667HG-B Super IO Sensors" (confidence: 9)
Note: there is no driver for [this sensor] yet.
Check http://www.lm-sensors.org/wiki/Devices for updates.

The link tells me that this chip is supported from kernel 2.6.36 (while Ubuntu 10.10 is using 2.6.35) or by using a standalone driver.

What is the simplest way to update lm-sensors?

Running PowerTOP on an Ubuntu 10.10 system, one of the suggestions are

Enable the CONFIG_PM_ADVANCED_DEBUG kernel configuration option. This option will allow PowerTOP to collect runtime power management statistics.

What is a kernel configuration option and how do I enable it?

Can I put it as a startup parameter somewhere, or would I have to compile a new kernel?

Samba gives me 24 MB/s read and 44 MB/s write, while ftp gives 97 and 112 MB/s under the same circumstances.

The documentation says that

Generally, you should find that Samba performs similarly to ftp at raw transfer speed.

In my case it clearly doesn't.

Where can I find tips on how to debug Samba performance?

Or alternatively tips for replacing Samba with something else? (I can't use ftp, unfortunately, as I need something that can be used with rsync/rsnapshot.)

More details:

• Both computers are running Ubuntu 10.10 (using Samba because I have a Mac as well)

• The Samba share is on a local home network, mounted as

  $ mount
  ...
  //server.local/share/ on /mnt/share type cifs (rw,mand)
  

• Samba performance was tested by copying (cp) a single file of ~4GB to and from the share, using time for timing and calculating transfer speed by hand.

• ftp performance are the numbers from the ftp client for get/put of the same file.
• iperf gives network speed ~900 Mbits/s
• bonnie++ gives disk speeds >200 MB/s on both sides for block reads as well as block writes
• Tried changing the parameters suggested in the performance tuning HOWTO (read/write raw, read size, socket options), most of them made little to no difference. (The one that made a difference caused write speed to drop 50%.)

I'm trying to create a bootable USB image to install Ubuntu on a new computer.

I have done this before following the "create USB drive" instructions for Ubuntu desktop, but I don't have an Ubuntu desktop available.

How can I do the same using only the command line?

Things I've tried:

• Create bootable USB on Mac OS X following the ubuntu.com "create USB drive" instructions for Mac: Doesn't boot.
• usb-creator: According to apt-cache search usb-creator and Wikipedia usb-creator only exists as a graphical tool.
• "Create manually" instructions at help.ubuntu.com: None of the files and directories described (e.g. casper, filesystem.manifest, menu.lst) exist in the ISO image, and I don't know what has replaced them.
• unetbootin scripting: Requires X server (graphics support) to run, even when fully scripted. (The command sudo unetbootin lang=en method=diskimage isofile=~/ubuntu-10.10-server-amd64.iso installtype=USB targetdrive=/dev/sdg1 autoinstall=yes gives an error message unetbootin: cannot connect to X server.)

Update

Also tried GRUB fiddling: Merging information from

• pendrivelinux.com
• a related question on the Linux Stackexchange
• and a grub configuration example

I was able to get halfway there - it booted from USB, displayed the grub menu and started the installation, but the installation did not complete.

For reference, this is the closest I got:

sudo su
  # mount USB pen
mount /dev/sd[X]1 /media/usb
  # install GRUB
grub-install --force --no-floppy --root-directory=/media/usb /dev/sd[X]
  # copy ISO image to USB
cp ~/ubuntu-10.10-server-amd64.iso /media/usb
  # mount ISO image, copy existing grub.cfg
mount ~/ubuntu-10.10-server-amd64.iso /media/iso/ -o loop
cp /media/iso/boot/grub/grub.cfg /media/usb/boot/grub/

I then edited /media/usb/boot/grub.cfg to add an .iso loopback, example grub entry:

menuentry "Install Ubuntu Server" {
  set gfxpayload=keep
  loopback loop /ubuntu-10.10-server-amd64.iso  
  linux (loop)/install/vmlinuz  file=(loop)/preseed/ubuntu-server.seed iso-scan/filename=/ubuntu-10.10-server-amd64.iso quiet --
  initrd (loop)/install/initrd.gz
}

When booting from USB, this would give me the Grub boot menu and start the installer, but the installer gave up after a couple of screens complaining that it couldn't find the CD-ROM drive. (Naturally, as the box I'm installing on doesn't have an optical drive.)

I resolved this particular issue by giving up and doing the "create USB drive" routine using the Ubuntu Live desktop CD (on a computer that does have an optical drive), then the USB install works.

But I expect that there is some way to do this from the command line of an Ubuntu system without X server and without an optical drive, so the question still stands.

Does anyone know how?

Occasionally I want to cd into a directory where my user does not have permission, so I resort to sudo.

The obvious command sudo cd somedir doesn't work:

$ sudo mkdir test
$ sudo chmod go-rxw test
$ ls -l
drwx------ 2 root     root  [...snip...] test
$ cd test
-bash: cd: test: Permission denied
$ sudo cd test
sudo: cd: command not found

Using sudo su works:

$ sudo su
# cd test

Is it possible to make this into a one-liner? (Not a big deal, just idle curiosity :)

The variations I tried didn't work:

$ sudo "cd test"
sudo: cd: command not found
$ sudo -i cd test
-bash: line 0: cd: test: No such file or directory
$ sudo -s cd test

The last one doesn't give an error, but it cd's within a new shell that exits by the end of the line, so it doesn't actually take me anywhere.

Can someone enlighten me as to why this happens? Why is sudo cd not found, when for example sudo ls ... works fine?

Is there a command that will show which file system (ext3, ext4, FAT32, ...) the various partitions and disks are using?

Similar to how sudo fdisk -l lists information about disks and partitions?

When creating a file system ( mkfs ...) the file system reserves 5% of the space for its own use because, according to man tune2fs:

Reserving some number of filesystem blocks for use by privileged processes is done to avoid filesystem fragmentation, and to allow system daemons, such as syslogd(8), to continue to function correctly after non-privileged processes are prevented from writing to the filesystem.

But with large drives 5% is quite a lot of space.

I have 4x1.5 TB drives for data storage (the OS runs on a separate disk), so the default setting would reserve 300 GB, which is an order of magnitude more than the the entire OS drive.

The reserved space can be tweaked, but what is a reasonable size for a data disk? Can I set it to zero, or could that lead to issues with fragmentation?

Are there any guidelines for which storage scheme(s) makes most sense for a multiple-disk home server?

I am assuming a separate boot/OS disk (so bootability is not a concern, this is for data storage only) and 4-6 storage disks of 1-2 TB each, for a total storage capacity in the range 4-12 TB.

The file system is ext4, I expect there will be only one big partition spanning all disks.

As far as I can tell, the alternatives are

individual disks

• pros: works with any combination of disk sizes; losing a disk loses only the data on that disk; no need for volume management.
• cons: data management is clumsy when logical units (like a "movies" folder) are larger than the capacity of any single drive.

JBOD span

• pros: can merge disks of any size.
• cons: losing a disk loses all data on all disks

LVM

• pros: can merge disks of any size; relatively simple to add and remove disks.
• cons: losing a disk loses all data on all disks

RAID 0

• pros: speed
• cons: losing one drive loses all data; disks must be same size

RAID 5

• pros: data survives losing one disk
• cons: gives up one disk worth of capacity; disks must be same size

RAID 6

• pros: data survives losing two disks
• cons: gives up two disks worth of capacity; disks must be same size

I'm primarily considering either LVM or JBOD span simply because it will let me reuse older, smaller-capacity disks when I upgrade the system. The runner-up is RAID 0 for speed.

I'm planning on having full backups to a separate system, so I expect the extra redundancy from RAID levels 5 or 6 won't be important.

Is this a fair representation of the alternatives? Are there other considerations or alternatives I have missed? And what would you recommend?