Posts Tagged ‘sector’

How to detect failing storage LUN on Linux – multipath

Wednesday, April 16th, 2014

detect-failing-lun-on-linux-failing-scsi-detection

If you login to server and after running dmesg – show kernel log command you get tons of messages like:

# dmesg

end_request: I/O error, dev sdc, sector 0
sd 0:0:0:1: SCSI error: return code = 0x00010000
end_request: I/O error, dev sdb, sector 0
sd 0:0:0:0: SCSI error: return code = 0x00010000
end_request: I/O error, dev sda, sector 0
sd 0:0:0:2: SCSI error: return code = 0x00010000
end_request: I/O error, dev sdc, sector 0
sd 0:0:0:1: SCSI error: return code = 0x00010000
end_request: I/O error, dev sdb, sector 0
sd 0:0:0:0: SCSI error: return code = 0x00010000
end_request: I/O error, dev sda, sector 0
sd 0:0:0:2: SCSI error: return code = 0x00010000
end_request: I/O error, dev sdc, sector 0
sd 0:0:0:1: SCSI error: return code = 0x00010000
end_request: I/O error, dev sdb, sector 0
sd 0:0:0:0: SCSI error: return code = 0x00010000
end_request: I/O error, dev sda, sector 0

In /var/log/messages there are also log messages present like:

# cat /var/log/messages
...

Apr 13 09:45:49 sl02785 kernel: end_request: I/O error, dev sda, sector 0
Apr 13 09:45:49 sl02785 kernel: klogd 1.4.1, ———- state change ———-
Apr 13 09:45:50 sl02785 kernel: sd 0:0:0:1: SCSI error: return code = 0x00010000
Apr 13 09:45:50 sl02785 kernel: end_request: I/O error, dev sdb, sector 0
Apr 13 09:45:55 sl02785 kernel: sd 0:0:0:2: SCSI error: return code = 0x00010000
Apr 13 09:45:55 sl02785 kernel: end_request: I/O error, dev sdc, sector 0

This is a sure sign something is wrong with SCSI hard drives or SCSI controller, to further debug the situation you can use the multipath command, i.e.:


multipath -ll | grep -i failed
_ 0:0:0:2 sdc 8:32 [failed][faulty]
_ 0:0:0:1 sdb 8:16 [failed][faulty]
_ 0:0:0:0 sda 8:0 [failed][faulty]

 

As you can see all 3 drives merged (sdc, sdb and sda) to show up on 1 physical drive via the remote network connectedLUN to the server is showing as faulty. This is a clear sign something is either wrong with 3 hard drive members of LUN – (Logical Unit Number) (which is less probable)  or most likely there is problem with the LUN network  SCSI controller. It is common error that LUN SCSI controller optics cable gets dirty and needs a physical clean up to solve it.

In case you don't know what is storage LUN? – In computer storage, a logical unit number, or LUN, is a number used to identify a logical unit, which is a device addressed by the SCSI protocol or protocols which encapsulate SCSI, such as Fibre Channel or iSCSI. A LUN may be used with any device which supports read/write operations, such as a tape drive, but is most often used to refer to a logical disk as created on a SAN. Though not technically correct, the term "LUN" is often also used to refer to the logical disk itself.

What LUN's does is very similar to Linux's Software LVM (Logical Volume Manager).

diskinfo Linux hdparm FreeBSD equivalent command for disk info and benchmarking

Thursday, March 8th, 2012

FreeBSD Linux hdparm equivalent is diskinfo artistic logo

On Linux there is the hdparm tool for various hard disk benchmarking and extraction of hard disk operations info.
As the Linux manual states hdparmget/set SATA/IDE device parameters

Most Linux users should already know it and might wonder if there is hdparm port or equivalent for FreeBSD, the aim of this short post is to shed some light on that.

The typical use of hdparm is like this:

linux:~# hdparm -t /dev/sda8

/dev/sda8:
Timing buffered disk reads: 76 MB in 3.03 seconds = 25.12 MB/sec
linux:~# hdparm -T /dev/sda8
/dev/sda8:
Timing cached reads: 1618 MB in 2.00 seconds = 809.49 MB/sec

The above output here is from my notebook Lenovo R61i.
If you're looking for alternative command to hdparm you should know in FreeBSD / OpenBSD / NetBSD, there is no exact hdparm equivalent command.
The somehow similar hdparm equivallent command for BSDs (FreeBSD etc.) is:
diskinfo

diskinfo is not so feature rich as linux's hdparm. It is just a simple command to show basic information for hard disk operations without no possibility to tune any hdd I/O and seek operations.
All diskinfo does is to show statistics for a hard drive seek times I/O overheads. The command takes only 3 arguments.

The most basic and classical use of the command is:

freebsd# diskinfo -t /dev/ad0s1a
/dev/ad0s1a
512 # sectorsize
20971520000 # mediasize in bytes (20G)
40960000 # mediasize in sectors
40634 # Cylinders according to firmware.
16 # Heads according to firmware.
63 # Sectors according to firmware.
ad:4JV48BXJs0s0 # Disk ident.

Seek times:
Full stroke: 250 iter in 3.272735 sec = 13.091 msec
Half stroke: 250 iter in 3.507849 sec = 14.031 msec
Quarter stroke: 500 iter in 9.705555 sec = 19.411 msec
Short forward: 400 iter in 2.605652 sec = 6.514 msec
Short backward: 400 iter in 4.333490 sec = 10.834 msec
Seq outer: 2048 iter in 1.150611 sec = 0.562 msec
Seq inner: 2048 iter in 0.215104 sec = 0.105 msec

Transfer rates:
outside: 102400 kbytes in 3.056943 sec = 33498 kbytes/sec
middle: 102400 kbytes in 2.696326 sec = 37978 kbytes/sec
inside: 102400 kbytes in 3.178711 sec = 32214 kbytes/sec

Another common use of diskinfo is to measure hdd I/O command overheads with -c argument:

freebsd# diskinfo -c /dev/ad0s1e
/dev/ad0s1e
512 # sectorsize
39112312320 # mediasize in bytes (36G)
76391235 # mediasize in sectors
75784 # Cylinders according to firmware.
16 # Heads according to firmware.
63 # Sectors according to firmware.
ad:4JV48BXJs0s4 # Disk ident.

I/O command overhead:
time to read 10MB block 1.828021 sec = 0.089 msec/sector
time to read 20480 sectors 4.435214 sec = 0.217 msec/sector
calculated command overhead = 0.127 msec/sector

Above diskinfo output is from my FreeBSD home router.

As you can see, the time to read 10MB block on my hard drive is 1.828021 (which is very high number),
this is a sign the hard disk experience too many read/writes and therefore needs to be shortly replaced with newer faster one.
diskinfo is part of the basis bsd install (bsd world). So it can be used without installing any bsd ports or binary packages.

For the purpose of stress testing hdd, or just some more detailed benchmarking on FreeBSD there are plenty of other tools as well.
Just to name a few:
 

  • rawio – obsolete in FreeBSD 7.x version branch (not available in BSD 7.2 and higher)
  • iozone, iozone21 – Tools to test the speed of sequential I/O to files
  • bonnie++ – benchmark tool capable of performing number of simple fs tests
  • bonnie – predecessor filesystem benchmark tool to bonnie++
  • raidtest – test performance of storage devices
  • mdtest – Software to test metadata performance on filsystems
  • filebench – tool for micro-benchmarking storage subsystems

Linux hdparm allows also changing / setting various hdd ATA and SATA settings. Similarly, to set and change ATA / SATA settings on FreeBSD there is the:

  • ataidle

tool.

As of time of writting ataidle is in port path /usr/ports/sysutils/ataidle/

To check it out install it as usual from the port location:

FreeBSD also has also the spindown port – a small program for handling automated spinning down ofSCSI harddrive
spindown is useful in setting values to SATA drives which has problems with properly controlling HDD power management.

To keep constant track on hard disk operations and preliminary warning in case of failing hard disks on FreeBSD there is also smartd service, just like in Linux.
smartd enables you to to control and monitor storage systems using the Self-Monitoring, Analysisand Reporting Technology System (S.M.A.R.T.) built into most modern ATA and SCSI hard disks.
smartd and smartctl are installable via the port /usr/ports/sysutils/smartmontools.

To install and use smartd, ataidle and spindown run:

freebsd# cd /usr/ports/sysutils/smartmontools
freebsd# make && make install clean
freebsd# cd /usr/ports/sysutils/ataidle/
freebsd# make && make install clean
freebsd# cd /usr/ports/sysutils/spindown/
freebsd# make && make install clean

Check each one's manual for more info.

How to check any filesystem for bad blocks using GNU / Linux or FreeBSD with dd

Monday, November 28th, 2011

Check any filesystem partition for BAD BLOCKS with DD on GNU Linux and FreeBSD

Have you looked for a universal physical check up tool to check up any filesystem type existing on your hard drive partitions?
I did! and was more than happy to just recently find out that the small UNIX program dd is capable to check any file system which is red by the Linux or *BSD kernel.

I’ll give an example, I have few partitions on my laptop computer with linux ext3 filesystem and NTFS partition.
My partitions looks like so:

noah:/home/hipo# fdisk -l
Disk /dev/sda: 160.0 GB, 160041885696 bytes
255 heads, 63 sectors/track, 19457 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x2d92834c
Device Boot Start End Blocks Id System
/dev/sda1 1 721 5786624 27 Unknown
Partition 1 does not end on cylinder boundary.
/dev/sda2 * 721 9839 73237024 7 HPFS/NTFS
/dev/sda3 9839 19457 77263200 5 Extended
/dev/sda5 9839 12474 21167968+ 83 Linux
/dev/sda6 12474 16407 31593208+ 83 Linux
/dev/sda7 16407 16650 1950448+ 82 Linux swap / Solaris
/dev/sda8 16650 19457 22551448+ 83 Linux

For all those unfamiliar with dddd – convert and copy a file this tiny program is capable of copying data from (if) input file to an output file as in UNIX , the basic philosophy is that everything is a file partitions themselves are also files.
The most common use of dd is to make image copies of a partition with any type of filesystem on it and move it to another system
Looking from a Windows user perspective dd is the command line Norton Ghost equivalent for Linux and BSD systems.
The classic way dd is used to copy let’s say my /dev/sda1 partition to another hard drive /dev/hdc1 is by cmds:

noah:/home/hipo# dd if=/dev/sda1 of=/dev/hdc1 bs=16065b

Even though the basic use of dd is to copy files, its flexibility allows a “trick” through which dd can be used to check any partition readable by the operating system kernel for bad blocks

In order to check any of the partitions listed, let’s say the one listed with filesystem HPFS/NTFS on /dev/sda2 using dd

noah:/home/hipo# dd if=/dev/sda2 of=/dev/null bs=1M

As you can see the of (output file) for dd is set to /dev/null in order to prevent dd to write out any output red by /dev/sda2 partition. bs=1M instructs dd to read from /dev/sda2 by chunks of 1 Megabyte in order to accelerate the speed of checking the whole drive.
Decreasing the bs=1M to less will take more time but will make the bad block checking be more precise.
Anyhow in most cases bs of 1 Megabyte will be a good value.

After some minutes (depending on the partition size), dd if, of operations outputs a statistics informing on how dd operations went.
Hence ff some of the blocks on the partition failed to be red by dd this will be shown in the final stats on its operation completion.
The drive, I’m checking does not have any bad blocks and dd statistics for my checked partition does not show any hard drive bad block problems:

71520+1 records in
71520+1 records out
74994712576 bytes (75 GB) copied, 1964.75 s, 38.2 MB/s

The statistics is quite self explanatory my partition of s size 75 GB was scanned for 1964 seconds roughly 32 minutes 46 seconds. The number of records red and written are 71520+1 e.g. (records in / records out). This means that all the records were properly red and wrote to /dev/null and therefore no BAD blocks on my NTFS partition 😉

How to reboot remotely Linux server if reboot, shutdown and init commands are not working (/sbin/reboot: Input/output error) – Reboot Linux in emergency using MagicSysRQ kernel sysctl variable

Saturday, July 23rd, 2011

SysRQ an alternative way to restart unrestartable Linux server

I’ve been in a situation today, where one Linux server’s hard drive SCSI driver or the physical drive is starting to break off where in dmesg kernel log, I can see a lot of errors like:

[178071.998440] sd 0:0:0:0: [sda] Result: hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK,SUGGEST_OK
[178071.998440] end_request: I/O error, dev sda, sector 89615868

I tried a number of things to remount the hdd which was throwing out errors in read only mode, but almost all commands I typed on the server were either shown as missng or returning an error:
Input/output error

Just ot give you an idea what I mean, here is a paste from the shell:

linux-server:/# vim /etc/fstab
-bash: vim: command not found
linux-server:/# vi /etc/fstab
-bash: vi: command not found
linux-server:/# mcedit /etc/fstab
-bash: /usr/bin/mcedit: Input/output error
linux-server:/# fdisk -l
-bash: /sbin/fdisk: Input/output error

After I’ve tried all kind of things to try to diagnose the server and all seemed failing, I thought next a reboot might help as on server boot the filesystems will get checked with fsck and fsck might be able to fix (at least temporary) the mess.

I went on and tried to restart the system, and guess what? I got:

/sbin/reboot init Input/output error

I hoped that at least /sbin/shutdown or /sbin/init commands might work out and since I couldn’t use the reboot command I tried this two as well just to get once again:

linux-server:/# shutdown -r now
bash: /sbin/shutdown: Input/output error
linux-server:/# init 6
bash: /sbin/init: Input/output error

You see now the situation was not pinky, it seemed there was no way to reboot the system …
Moreover the server is located in remote Data Center and I the tech support there is conducting assigned task with the speed of a turtle.
The server had no remote reboot, web front end or anything and thefore I needed desperately a way to be able to restart the machine.

A bit of research on the issue has led me to other people who experienced the /sbin/reboot init Input/output error error mostly caused by servers with failing hard drives as well as due to HDD control driver bugs in the Linux kernel.

As I was looking for another alternative way to reboot my Linux machine in hope this would help. I came across a blog post Rebooting the Magic Wayhttp://www.linuxjournal.com/content/rebooting-magic-way

As it was suggested in Cory’s blog a nice alternative way to restart a Linux machine without using reboot, shutdown or init cmds is through a reboot with the Magic SysRQ key combination

The only condition for the Magic SysRQ key to work is to have enabled the SysRQ – CONFIG_MAGIC_SYSRQ in Kernel compile time.
As of today luckily SysRQ Magic key is compiled and enabled by default in almost all modern day Linux distributions in this numbers Debian, Fedora and their derivative distributions.

To use the sysrq kernel capabilities as a mean to restart the server, it’s necessery first to activate the sysrq through sysctl, like so:

linux-server:~# sysctl -w kernel.sysrq=1
kernel.sysrq = 1

I found enabling the kernel.sysrq = 1 permanently in the kernel is also quite a good idea, to achieve that I used:

echo 'kernel.sysrq = 1' >> /etc/sysctl.conf

Next it’s wise to use the sync command to sync any opened files on the server as well stopping as much of the server active running services (MySQL, Apache etc.).

linux-server:~# sync

Now to reboot the Linux server, I used the /proc Linux virtual filesystem by issuing:

linux-server:~# echo b > /proc/sysrq-trigger

Using the echo b > /proc/sysrq-trigger simulates a keyboard key press which does invoke the Magic SysRQ kernel capabilities and hence instructs the kernel to immediately reboot the system.
However one should be careful with using the sysrq-trigger because it’s not a complete substitute for /sbin/reboot or /sbin/shutdown -r commands.
One major difference between the standard way to reboot via /sbin/reboot is that reboot kills all the running processes on the Linux machine and attempts to unmount all filesystems, before it proceeds to sending the kernel reboot instruction.

Using echo b > /proc/sysrq-trigger, however neither tries to umount mounted filesystems nor tries to kill all processes and sync the filesystem, so on a heavy loaded (SQL data critical) server, its use might create enormous problems and lead to severe data loss!

SO BEWARE be sure you know what you’re doing before you proceed using /proc/sysrq-trigger as a way to reboot ;).