Linux: Difference between revisions

From DWIKI
mNo edit summary
 
(25 intermediate revisions by the same user not shown)
Line 8: Line 8:
*[[Distributions|Distributions]]  
*[[Distributions|Distributions]]  
*https://kb.novaordis.com/index.php/Events_OS_Metrics  
*https://kb.novaordis.com/index.php/Events_OS_Metrics  
 
*Linux on ircnet: #linux2
 
 


Line 20: Line 20:
 
 


== Checking resources ==
*inxi
*[http://www.slashroot.in/linux-system-io-monitoring http://www.slashroot.in/linux-system-io-monitoring]
=== show what is doing most disk accesses ===
iotop
glances
=== IO statistics ===
iostat
vmstat
dstat
ioping
atop
=== CPU usage etc ===
top
atop
htop
vtop
fio
pidstat


== Administration ==
== Administration ==
Line 62: Line 36:




== FAQ ==
= FAQ =
==Storage==
===Umount: /mnt/sda1 is busy===
Try
umount -lf /mnt/sda1
 
=== Grow (GPT) last partition to max available ===
 
Where '3' is partition number
 
parted /dev/sdf resize 3 100%
 
No longer allowed?
 
gdisk
 
===I deleted a file but disk is still full===
It was in use by process, you'll need to kill that
lsof -nP | grep '(deleted)'
etc
 
 
===Which disk is 0.0.3.0===
lsblk -dno name,hctl
 
 
===Power-on or device reset occurred===
I had a case where megamaid decided to spin down "unused disks" that were in JBOD. Disable that option ot switch the controller to '''IT mode'''.
 
==Network==


=== check if interface exists ===
=== check if interface exists ===
Line 68: Line 71:
/sys/class/net/$IF
/sys/class/net/$IF


=== check if virtual of physicalmachine ===
dmidecode -s system-manufacturer
virt-who


 
==Hardware info==
===Check number of CPUs===
lscpu | grep Socket


===Find bios version===
===Find bios version===
  dmidecode -s bios-version
  dmidecode -s bios-version


=== watch which files a process opens ===
watch ls -l /proc/`pidof clamd`/fd


=== pam_succeed_if(sudo:auth): requirement "uid >= 1000" not met ===
=== pam_succeed_if(sudo:auth): requirement "uid >= 1000" not met ===


=== Grow (GPT) last partition to max available ===
Where '3' is partition number
parted /dev/sdf resize 3 100%
No longer allowed?


gdisk


=== timestamp to human readable ===


date -d @1522142497


=== List hardware ===
=== List hardware ===
Line 104: Line 92:
*lsusb  
*lsusb  
*hwinfo  
*hwinfo  
*lspci  
*lspci
*lscpu


  dmidecode -t baseboard
  dmidecode -t baseboard
Line 115: Line 104:


=== calling ioclt to re-read partition table: Device or resource busy ===
=== calling ioclt to re-read partition table: Device or resource busy ===


=== rescan partition table ===
=== rescan partition table ===
Line 127: Line 117:
  kpartx /dev/sdg
  kpartx /dev/sdg
  partx -uv /dev/sdg (this one worked on centos 7.3!)
  partx -uv /dev/sdg (this one worked on centos 7.3!)
This one gave new disk size!
Verified: CentOS 6, Ubuntu 22
  echo 1 > /sys/block/sde/device/rescan
  echo 1 > /sys/block/sde/device/rescan


Line 132: Line 125:


  echo "- - -" > /sys/class/scsi_host/host2/scan
  echo "- - -" > /sys/class/scsi_host/host2/scan
or scan all:
echo "- - -" | tee /sys/class/scsi_host/host*/scan


The winner so far:
The winner so far:


  blockdev --rereadpt /dev/sdg
  blockdev --rereadpt /dev/sdg
but may throw blockdev: ioctl error on BLKRRPART: Device or resource busy


maybe
maybe
Line 151: Line 148:
  #reboot
  #reboot
  echo b > /proc/sysrq-trigger
  echo b > /proc/sysrq-trigger
==Misc==
===check if you're in virtual or physical machine===
dmidecode -s system-manufacturer
virt-who
lscpu
virt-what
=== watch which files a process opens ===
watch ls -l /proc/`pidof clamd`/fd
=== timestamp to human readable ===
date -d @1522142497


=== Mignight commander/dialog strange characters ===
=== Mignight commander/dialog strange characters ===
Line 200: Line 214:


try kpartx instead
try kpartx instead
===Find maximum depth of directory===
find /var/www -type d | awk -F"/" 'NF > max {max = NF} END {print max}'
   [[Category:System Administration]] [[Category:Linux]]
   [[Category:System Administration]] [[Category:Linux]]
==Booting==
===Cannot open access to console, the root account is locked===
You probably messed up your grub config, maybe GRUB_CMDLINE_LINUX. Boot rescue CD image :)

Latest revision as of 08:10, 25 July 2024

Linux is a free Unix-type operating system originally created by Linus Torvalds with the assistance of developers around the world. Developed under the GNU General Public License , the source code for Linux is freely available to everyone.


Links

 

Rescue CDs

 


Administration

 



FAQ

Storage

Umount: /mnt/sda1 is busy

Try 

umount -lf /mnt/sda1

Grow (GPT) last partition to max available

Where '3' is partition number 

parted /dev/sdf resize 3 100%

No longer allowed? 

gdisk

I deleted a file but disk is still full

It was in use by process, you'll need to kill that 

lsof -nP | grep '(deleted)'

etc


Which disk is 0.0.3.0

lsblk -dno name,hctl


Power-on or device reset occurred

I had a case where megamaid decided to spin down "unused disks" that were in JBOD. Disable that option ot switch the controller to IT mode.

Network

check if interface exists

/sys/class/net/$IF


Hardware info

Check number of CPUs

lscpu | grep Socket

Find bios version

dmidecode -s bios-version


pam_succeed_if(sudo:auth): requirement "uid >= 1000" not met

List hardware

  • lshw
  • dmidecode
  • lsusb
  • hwinfo
  • lspci
  • lscpu
dmidecode -t baseboard

BLKRRPART: Device or resource busy

spend all day or reboot

 

calling ioclt to re-read partition table: Device or resource busy

rescan partition table

Keep an eye on 

cat /proc/partitions

Partition(s) have been written, but we have been unable to inform the kernel of the change 

partprobe
kpartx /dev/sdg
partx -uv /dev/sdg (this one worked on centos 7.3!)

This one gave new disk size! Verified: CentOS 6, Ubuntu 22 

echo 1 > /sys/block/sde/device/rescan

Only one that seemed to work on (centos7.x) 

echo "- - -" > /sys/class/scsi_host/host2/scan

or scan all: 

echo "- - -" | tee /sys/class/scsi_host/host*/scan

The winner so far: 

blockdev --rereadpt /dev/sdg

but may throw blockdev: ioctl error on BLKRRPART: Device or resource busy


maybe 

hdparm -z /dev/sdg

Force reboot

I found sometimes 'reboot' and 'shutdown' don't work in virtual machines, in that case try: 

#sync
echo s > /proc/sysrq-trigger
#optionally umount
echo u > /proc/sysrq-trigger
#reboot
echo b > /proc/sysrq-trigger

Misc

check if you're in virtual or physical machine

dmidecode -s system-manufacturer
virt-who
lscpu
virt-what

watch which files a process opens

watch ls -l /proc/`pidof clamd`/fd

timestamp to human readable

date -d @1522142497

Mignight commander/dialog strange characters

Quick fix: 

export LANG=en_US.ISO-8859-1

This probleem seems to be related to screen/su -

kpartx failing silently

No output/result when trying to create: 

kpartx -lv /dev/mapper/foo

this might means the device hasn't been partitioned

No output when trying to delete: 

kpartx -d /dev/mapper/foop1

means you should use 

kpartx -d /dev/mapper/foo

 

 

 

Create swapfile

fallocate -l 4G /var/swapfile
mkswap /var/swapfile

Temporary: 

swapon -v /var/swapfile

In fstab: /var/swapfile swap swap sw,prio=-1

swapon failed: Invalid argument 

problably trying to use fallocate on rhel/centos, use dd instead

partx: error adding partition

try kpartx instead


Find maximum depth of directory

find /var/www -type d | awk -F"/" 'NF > max {max = NF} END {print max}'



Booting

Cannot open access to console, the root account is locked

You probably messed up your grub config, maybe GRUB_CMDLINE_LINUX. Boot rescue CD image :)

Retrieved from "https://wiki.dhits.nl/index.php?title=Linux&oldid=9051"