Author: mirek

Linuxe

NVIDIA driver installation in KALI Linux

Written by mirek Published Discussion Leave a Comment on NVIDIA driver installation in KALI Linux

My old notebook HP EliteBook 8560w is not suitable for Win10 and I have decided to use it for Linux tests. The first Linux OS, I have installed, was Ubuntu. It enabled graphic card NVIDIA GF106GLM [Quadro 2000M] without any problems in GNOME. As the second OS I wanted to use KALI Linux as dual boot on the notebook. I was surprised immediately after the installation. Xfce did not work and I have to start the Linux configuration via text console (CTRL+ALT+F1). The first link with NVIDIA drivers description I have found here. There is a description of NVIDIA cards identification and a procedure for driver installation. Here are details for NVIDIA graphic card in HP EliteBook 8560w:

# lspci|grep -i vga
01:00.0 VGA compatible controller: NVIDIA Corporation GF106GLM [Quadro 2000M] (rev a1)

# lspci -s 01:00.0 -v
01:00.0 VGA compatible controller: NVIDIA Corporation GF106GLM [Quadro 2000M] (rev a1) (prog-if 00 [VGA controller])
DeviceName: 0
Subsystem: Hewlett-Packard Company GF106GLM [Quadro 2000M]
Flags: bus master, fast devsel, latency 0, IRQ 16, IOMMU group 1
Memory at d0000000 (32-bit, non-prefetchable) [size=32M]
Memory at c0000000 (64-bit, prefetchable) [size=128M]
Memory at c8000000 (64-bit, prefetchable) [size=64M]
I/O ports at 4000 [size=128]
Expansion ROM at 000c0000 [virtual] [disabled] [size=128K]
Capabilities: [60] Power Management version 3
Capabilities: [68] MSI: Enable- Count=1/1 Maskable- 64bit+
Capabilities: [78] Express Endpoint, MSI 00
Capabilities: [b4] Vendor Specific Information: Len=14 <?>
Capabilities: [100] Virtual Channel
Capabilities: [128] Power Budgeting <?>
Capabilities: [600] Vendor Specific Information: ID=0001 Rev=1 Len=024 <?>
Kernel driver in use: nouveau
Kernel modules: nouveau

KALI Linux used “nouveau” driver (Ubuntu used same), but GUI did not run. I tried to install the driver according to procedure here, but driver version 430.xx was installed and it is not suitable for my card. I have found the supported driver on this page and it should be version 390.xx (legacy driver). Here is link for KALI Linux installation, but the procedure did not install the driver properly. The card was still using the “nouveau” driver and GUI did not start. After several reinstallations of the driver and even KALI Linux installation I have found following procedure. It describes installation of NVIDIA driver downloaded directly from NVIDIA pages. It compiles kernel modules during the installation and it makes the proper GUI configuration. I have found info that the max. driver version is 390.132, but it is not true. After several tries to install driver 390.132 and 390.144 I had to use the latest driver 390.147, because of the error “fatal error: asm/kmap_types.h: No such file or directory” during kernel modules compilation. The error above has been fixed in the latest driver 390.147 based on the Change log for Ubuntu. The procedure to install the latest driver is as following:

Search for all NVIDIA drivers installed in the system

# dpkg -l|grep -i nvidia

Deinstall them

# apt-get remove –purge nvidia-*

Deinstall even libraries to be sure

# apt-get remove –purge libnvidia-*

Prevent loading of  “nouveau” driver

# sed ‘s/quiet/quiet nouveau.modeset=0/g’ -i /etc/default/grub
# update-grub
# reboot

Stop GUI

# service gdm3 stop

Download driver 390.147

# wget -c https://us.download.nvidia.com/XFree86/Linux-x86_64/390.147/NVIDIA-Linux-x86_64-390.147.run

Change the privileges

# chmod +x NVIDIA-Linux-x86_64-390.147.run

Install it. Register kernel modules with DKMS and refuse to install 32-bit drivers. There is an error reported during the installation about missing path to glvnd. It is possible to ignore it or fix by adding path parameter to the driver install command like below

# ./NVIDIA-Linux-x86_64-390.147.run –glvnd-egl-config-path=/etc/glvnd/egl_vendor.d

After reboot GUI starts without any problem

# reboot

The properly installed driver looks like this

# lspci|grep -i vga
01:00.0 VGA compatible controller: NVIDIA Corporation GF106GLM [Quadro 2000M] (rev a1)

# lspci -s 01:00.0 -v
01:00.0 VGA compatible controller: NVIDIA Corporation GF106GLM [Quadro 2000M] (rev a1) (prog-if 00 [VGA controller])
DeviceName: 0
Subsystem: Hewlett-Packard Company GF106GLM [Quadro 2000M]
Flags: bus master, fast devsel, latency 0, IRQ 44, IOMMU group 1
Memory at d0000000 (32-bit, non-prefetchable) [size=32M]
Memory at c0000000 (64-bit, prefetchable) [size=128M]
Memory at c8000000 (64-bit, prefetchable) [size=64M]
I/O ports at 4000 [size=128]
Expansion ROM at 000c0000 [virtual] [disabled] [size=128K]
Capabilities: [60] Power Management version 3
Capabilities: [68] MSI: Enable+ Count=1/1 Maskable- 64bit+
Capabilities: [78] Express Endpoint, MSI 00
Capabilities: [b4] Vendor Specific Information: Len=14 <?>
Capabilities: [100] Virtual Channel
Capabilities: [128] Power Budgeting <?>
Capabilities: [600] Vendor Specific Information: ID=0001 Rev=1 Len=024 <?>
Kernel driver in use: nvidia
Kernel modules: nouveau, nvidia_drm, nvidia


# nvidia-smi
Wed Apr 6 14:46:36 2022
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 390.147                                  Driver Version: 390.147 |
|-------------------------------+----------------------+----------------------+
| GPU Name         Persistence-M| Bus-Id Disp.A        | Volatile Uncorr. ECC |
| Fan Temp Perf    Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|===============================+======================+======================|
| 0 Quadro 2000M            Off | 00000000:01:00.0 N/A |                  N/A |
| N/A 28C  P12    N/A /     N/A |     123MiB / 1985MiB | N/A          Default |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes:                                                       GPU Memory |
| GPU         PID      Type Process name                                Usage |
|=============================================================================|
| 0                    Not Supported                                          |
+-----------------------------------------------------------------------------+

Continue reading NVIDIA driver installation in KALI Linux
Linuxe, Uncategorized

Shrink CentOS 8.1 XFS disk in VM

Written by mirek Published Discussion Leave a Comment on Shrink CentOS 8.1 XFS disk in VM

New version of OS (CentOS 8.1) was installed to new VM with following layout:

# lvs

LV   VG Attr       LSize   Pool Origin Data%  Meta%  Move Log Cpy%Sync Convert

home cl -wi-a—– 441.10g

root cl -wi-ao—-  50.00g

swap cl -wi-ao—-   7.89g

# pvdisplay /dev/sda2 -m

— Physical volume —

PV Name               /dev/sda2

VG Name               cl

PV Size               <499.00 GiB / not usable 3.00 MiB

Allocatable           yes

PE Size               4.00 MiB

Total PE              127743

Free PE               110362

Allocated PE          17381

PV UUID               fB3mro-9lDl-lMBg-0jNh-qqzx-PReV-iHUu8T

 

— Physical Segments —

Physical extent 0 to 2020:

Logical volume      /dev/cl/swap

Logical extents     0 to 2020

Physical extent 2021 to 114942:

Logical volume      /dev/cl/home

Logical extents     0 to 112921

Physical extent 114943 to 127742:

Logical volume      /dev/cl/root

Logical extents     0 to 12799

 

We have 440GB disk space for home directory. There is one user and it has nothing in his directory. We know that XFS can not be shrunk and we have to do following exercise:-)

# ll /home

total 0

drwx——. 2 xpadm xpadm 6 Jun 16 09:40 xpadm

# umount /home

 

Remove /home directory and we will recreate it with smaller size.

# lvremove /dev/cl/home

Do you really want to remove active logical volume cl/home? [y/n]: y

Logical volume “home” successfully removed

 

Our new size will be 10GB:

# lvcreate -L 10G -n home cl

WARNING: xfs signature detected on /dev/cl/home at offset 0. Wipe it? [y/n]: y

Wiping xfs signature on /dev/cl/home.

Logical volume “home” created.

 

The OS complains about existing XFS filesystem on volume, but it is OK we wanted to delete it

# vgchange -a y

3 logical volume(s) in volume group “cl” now active

Create a new XFS filesystem and restore home directory on it

# mkfs.xfs /dev/cl/home

meta-data=/dev/cl/home           isize=512    agcount=4, agsize=655360 blks

=                       sectsz=512   attr=2, projid32bit=1

=                       crc=1        finobt=1, sparse=1, rmapbt=0

=                       reflink=1

data     =                       bsize=4096   blocks=2621440, imaxpct=25

=                       sunit=0      swidth=0 blks

naming   =version 2              bsize=4096   ascii-ci=0, ftype=1

log      =internal log           bsize=4096   blocks=2560, version=2

=                       sectsz=512   sunit=0 blks, lazy-count=1

realtime =none                   extsz=4096   blocks=0, rtextents=0

# mount /home

# ll /home

total 0

# mkdir /home/xpadm

# ll /home

total 0

drwxr-xr-x. 2 root root 6 Jun 16 09:40 xpadm

# chmod 700 /home/xpadm/

# chown xpadm:xpadm /home/xpadm/

# ll /home

total 0

drwx——. 2 xpadm xpadm 6 Jun 16 09:40 xpadm

 

We have freed up plenty of space and we need to prepare vg cl for shrinking:

# vgs

VG #PV #LV #SN Attr   VSize    VFree

cl   1   3   0 wz–n- <499.00g 431.10g

# pvs

PV         VG Fmt  Attr PSize    PFree

/dev/sda2  cl lvm2 a–  <499.00g 431.10g

 

 

Let’s check details how physical segments are allocated inside vg cl

# pvdisplay /dev/sda2 -m

— Physical volume —

PV Name               /dev/sda2

VG Name               cl

PV Size               <499.00 GiB / not usable 3.00 MiB

Allocatable           yes

PE Size               4.00 MiB

Total PE              127743

Free PE               110362

Allocated PE          17381

PV UUID               fB3mro-9lDl-lMBg-0jNh-qqzx-PReV-iHUu8T

 

— Physical Segments —

Physical extent 0 to 2020:

Logical volume      /dev/cl/swap

Logical extents     0 to 2020

Physical extent 2021 to 4580:

Logical volume      /dev/cl/home

Logical extents     0 to 2559

Physical extent 4581 to 114942:

FREE

Physical extent 114943 to 127742:

Logical volume      /dev/cl/root

Logical extents     0 to 12799

 

We have free space between the new /home and the old /. If we shrink the space now we will cut the / and the system will be lost. We need to move the FREE segments to the end of the disk space. Let’s do a small trick by adding new disk /dev/sdb1 to vg and then move / physical extents to the new physical volume in vg cl. It is possible to do it online and it will free up the disk space up to the end of physical volume /dev/sda2. Then we will move them back and OS will use extents right behind our new smaller /home directory.

 

# pvcreate /dev/sdb1

Physical volume “/dev/sdb1” successfully created.

# vgextend cl /dev/sdb1

Volume group “cl” successfully extended

# vgs

VG #PV #LV #SN Attr   VSize   VFree

cl   2   3   0 wz–n- 648.99g <581.10g

# pvs -o+pv_used

PV         VG Fmt  Attr PSize    PFree    Used

/dev/sda2  cl lvm2 a–  <499.00g  431.10g 67.89g

/dev/sdb1  cl lvm2 a–  <150.00g <150.00g     0

# pvmove -n /dev/cl/root /dev/sda2 /dev/sdb1

/dev/sda2: Moved: 0.02%

/dev/sda2: Moved: 3.41%

…

/dev/sda2: Moved: 96.80%

/dev/sda2: Moved: 100.00%

 

# pvmove -n /dev/cl/root /dev/sdb1 /dev/sda2

/dev/sdb1: Moved: 0.00%

/dev/sdb1: Moved: 3.09%

…

/dev/sdb1: Moved: 98.88%

/dev/sdb1: Moved: 100.00%

 

When we check the space in physical volume /dev/sda2 we can see that extents were properly copied behind /home and there is plenty of space at the end of the physical volume from physical extent 17831.

# pvdisplay /dev/sda2 -m

— Physical volume —

PV Name               /dev/sda2

VG Name               cl

PV Size               <499.00 GiB / not usable 3.00 MiB

Allocatable           yes

PE Size               4.00 MiB

Total PE              127743

Free PE               110362

Allocated PE          17381

PV UUID               fB3mro-9lDl-lMBg-0jNh-qqzx-PReV-iHUu8T

 

— Physical Segments —

Physical extent 0 to 2020:

Logical volume      /dev/cl/swap

Logical extents     0 to 2020

Physical extent 2021 to 4580:

Logical volume      /dev/cl/home

Logical extents     0 to 2559

Physical extent 4581 to 17380:

Logical volume      /dev/cl/root

Logical extents     0 to 12799

Physical extent 17381 to 127742:

FREE

 

 

Now we have free space at the end of physical disk /dev/sda2 and we can shrink PV to 80GB. (Yes, we can do calculation here, but it will not be fun) The size of 80 GB is enough to accommodate the disk space for current swap, /home and /.

# fdisk -l /dev/sda

Disk /dev/sda: 500 GiB, 536870912000 bytes, 1048576000 sectors

Units: sectors of 1 * 512 = 512 bytes

Sector size (logical/physical): 512 bytes / 512 bytes

I/O size (minimum/optimal): 512 bytes / 512 bytes

Disklabel type: dos

Disk identifier: 0x845c40d9

 

Device     Boot   Start        End    Sectors  Size Id Type

/dev/sda1  *       2048    2099199    2097152    1G 83 Linux

/dev/sda2       2099200 1048575999 1046476800  499G 8e Linux LVM

# pvresize –setphysicalvolumesize 80G /dev/sda2

/dev/sda2: Requested size 80.00 GiB is less than real size <499.00 GiB. Proceed?  [y/n]: y

WARNING: /dev/sda2: Pretending size is 167772160 not 1046476800 sectors.

Physical volume “/dev/sda2” changed

1 physical volume(s) resized or updated / 0 physical volume(s) not resized

 

The command pvresize complains about our size, because it is smaller than actual size of /dev/sda2 (499GB). Anyway it works fine and our free space was decreased and there are still free extents for future use.

# pvdisplay /dev/sda2 -m

— Physical volume —

PV Name               /dev/sda2

VG Name               cl

PV Size               <80.00 GiB / not usable 3.00 MiB

Allocatable           yes

PE Size               4.00 MiB

Total PE              20479

Free PE               3098

Allocated PE          17381

PV UUID               fB3mro-9lDl-lMBg-0jNh-qqzx-PReV-iHUu8T

 

— Physical Segments —

Physical extent 0 to 2020:

Logical volume      /dev/cl/swap

Logical extents     0 to 2020

Physical extent 2021 to 4580:

Logical volume      /dev/cl/home

Logical extents     0 to 2559

Physical extent 4581 to 17380:

Logical volume      /dev/cl/root

Logical extents     0 to 12799

Physical extent 17381 to 20478:

FREE

 

 

We do not need additional disk in vg cl then we will remove it:

# vgreduce cl /dev/sdb1

Removed “/dev/sdb1” from volume group “cl”

# vgs

VG #PV #LV #SN Attr   VSize   VFree

cl   1   3   0 wz–n- <80.00g 12.10g

# vgdisplay

— Volume group —

VG Name               cl

System ID

Format                lvm2

Metadata Areas        1

Metadata Sequence No  15

VG Access             read/write

VG Status             resizable

MAX LV                0

Cur LV                3

Open LV               3

Max PV                0

Cur PV                1

Act PV                1

VG Size               <80.00 GiB

PE Size               4.00 MiB

Total PE              20479

Alloc PE / Size       17381 / 67.89 GiB

Free  PE / Size       3098 / 12.10 GiB

VG UUID               0XLSu8-aOeE-Mdg1-5ViQ-2zaQ-9hPv-X9T4h1

 

The new layout of vg cl reports size only 80 GB.

 

# vgdisplay -v

— Volume group —

VG Name               cl

System ID

Format                lvm2

Metadata Areas        1

Metadata Sequence No  15

VG Access             read/write

VG Status             resizable

MAX LV                0

Cur LV                3

Open LV               3

Max PV                0

Cur PV                1

Act PV                1

VG Size               <80.00 GiB

PE Size               4.00 MiB

Total PE              20479

Alloc PE / Size       17381 / 67.89 GiB

Free  PE / Size       3098 / 12.10 GiB

VG UUID               0XLSu8-aOeE-Mdg1-5ViQ-2zaQ-9hPv-X9T4h1

 

— Logical volume —

LV Path                /dev/cl/swap

LV Name                swap

VG Name                cl

LV UUID                3yJjj2-cLdt-uaiT-C3er-i2Zx-u0JY-Ip0xG4

LV Write Access        read/write

LV Creation host, time localhost, 2020-03-12 09:23:04 +0100

LV Status              available

# open                 2

LV Size                7.89 GiB

Current LE             2021

Segments               1

Allocation             inherit

Read ahead sectors     auto

– currently set to     8192

Block device           253:1

 

— Logical volume —

LV Path                /dev/cl/root

LV Name                root

VG Name                cl

LV UUID                LAmWX2-3DSS-VD7d-9rWs-SFzc-hTni-akto3r

LV Write Access        read/write

LV Creation host, time localhost, 2020-03-12 09:23:06 +0100

LV Status              available

# open                 1

LV Size                50.00 GiB

Current LE             12800

Segments               1

Allocation             inherit

Read ahead sectors     auto

– currently set to     8192

Block device           253:0

 

— Logical volume —

LV Path                /dev/cl/home

LV Name                home

VG Name                cl

LV UUID                3NCUe4-JZVY-FdkD-h8mC-Xuuz-LL6M-uslaFo

LV Write Access        read/write

LV Creation host, time host.example.net, 2020-06-16 09:39:03 +0200

LV Status              available

# open                 1

LV Size                10.00 GiB

Current LE             2560

Segments               1

Allocation             inherit

Read ahead sectors     auto

– currently set to     8192

Block device           253:2

 

— Physical volumes —

PV Name               /dev/sda2

PV UUID               fB3mro-9lDl-lMBg-0jNh-qqzx-PReV-iHUu8T

PV Status             allocatable

Total PE / Free PE    20479 / 3098

 

 

OK, the LVM part is done and we are ready to shrink the physical volume /dev/sda. It has 2 partitions. The 1st one is used for booting and the second as LVM physical volume for swap, /home and root.

We will dump the partition layout of /dev/sda to file via sfdisk command.

# sfdisk -d /dev/sda sda_part

label: dos

label-id: 0x845c40d9

device: /dev/sda

unit: sectors

 

/dev/sda1 : start=        2048, size=     2097152, type=83, bootable

/dev/sda2 : start=     2099200, size=  1046476800, type=8e

# sfdisk -d /dev/sda >sda_part

 

 

The editing of dump file we can shrink partition 2 of /dev/sda to the new size. The target was 20000 extents in LVM and each is 4MB. We have sectors 512 bytes then we need 20000 * 8192 = 163840000 sectors. Unfortunately it is only /dev/sda2 and I did not add /dev/sda1, but used following value for whole disk.

# vi sda_part

# cat sda_part

label: dos

label-id: 0x845c40d9

device: /dev/sda

unit: sectors

 

/dev/sda1 : start=        2048, size=     2097152, type=83, bootable

/dev/sda2 : start=     2099200, size=  163840000, type=8e

 

 

The file is applied by sfdisk command with option –no-reread, because we can not change it for running kernel. It was reported as Device or resource busy.

# sfdisk /dev/sda <sda_part –no-reread

Disk /dev/sda: 500 GiB, 536870912000 bytes, 1048576000 sectors

Units: sectors of 1 * 512 = 512 bytes

Sector size (logical/physical): 512 bytes / 512 bytes

I/O size (minimum/optimal): 512 bytes / 512 bytes

Disklabel type: dos

Disk identifier: 0x845c40d9

 

Old situation:

 

Device     Boot   Start        End    Sectors  Size Id Type

/dev/sda1  *       2048    2099199    2097152    1G 83 Linux

/dev/sda2       2099200 1048575999 1046476800  499G 8e Linux LVM

 

>>> Script header accepted.

>>> Script header accepted.

>>> Script header accepted.

>>> Script header accepted.

>>> Created a new DOS disk label with disk identifier 0x845c40d9.

/dev/sda1: Created a new partition 1 of type ‘Linux’ and of size 1 GiB.

Partition #1 contains a ext4 signature.

/dev/sda2: Created a new partition 2 of type ‘Linux LVM’ and of size 78.1 GiB.

Partition #2 contains a LVM2_member signature.

/dev/sda3: Done.

 

New situation:

Disklabel type: dos

Disk identifier: 0x845c40d9

 

Device     Boot   Start       End   Sectors  Size Id Type

/dev/sda1  *       2048   2099199   2097152    1G 83 Linux

/dev/sda2       2099200 165939199 163840000 78.1G 8e Linux LVM

 

The partition table has been altered.

Calling ioctl() to re-read partition table.

Re-reading the partition table failed.: Device or resource busy

The kernel still uses the old table. The new table will be used at the next reboot or after you run partprobe(8) or kpartx(8).

Syncing disks.

 

 

Then the command partprobe has fixed the issue, but I have done reboot to be sure that we can continue.

# partprobe

# fdisk -l /dev/sda

Disk /dev/sda: 500 GiB, 536870912000 bytes, 1048576000 sectors

Units: sectors of 1 * 512 = 512 bytes

Sector size (logical/physical): 512 bytes / 512 bytes

I/O size (minimum/optimal): 512 bytes / 512 bytes

Disklabel type: dos

Disk identifier: 0x845c40d9

 

Device     Boot   Start       End   Sectors  Size Id Type

/dev/sda1  *       2048   2099199   2097152    1G 83 Linux

/dev/sda2       2099200 165939199 163840000 78.1G 8e Linux LVM

# reboot

 

 

After reboot pvs is complaining that /dev/sda2 is smaller than PV created on it. We have 2 choices to fix it:

  1. Shrink PV
  2. Extend /dev/sda2

The option to extend /dev/sda2 needs dumping and editing of disk table, it is easier to shrink PV:

# pvs

WARNING: Device /dev/sda2 has size of 163840000 sectors which is smaller than corresponding PV size of 167770112 sectors. Was device resized?

WARNING: One or more devices used as PVs in VG cl have changed sizes.

PV         VG Fmt  Attr PSize    PFree

/dev/sda2  cl lvm2 a–   <80.00g   12.10g

# fdisk -l /dev/sda

Disk /dev/sda: 78.1 GiB, 83886080000 bytes, 163840000 sectors

Units: sectors of 1 * 512 = 512 bytes

Sector size (logical/physical): 512 bytes / 512 bytes

I/O size (minimum/optimal): 512 bytes / 512 bytes

Disklabel type: dos

Disk identifier: 0x845c40d9

 

Device     Boot   Start       End   Sectors  Size Id Type

/dev/sda1  *       2048   2099199   2097152    1G 83 Linux

/dev/sda2       2099200 165939199 163840000 78.1G 8e Linux LVM

 

 

After checking real size of /dev/sda2 we can do pvresize:

# pvresize –setphysicalvolumesize 161740800s /dev/sda2

WARNING: Device /dev/sda2 has size of 161740800 sectors which is smaller than corresponding PV size of 167770112 sectors. Was device resized?

WARNING: One or more devices used as PVs in VG cl have changed sizes.

Physical volume “/dev/sda2” changed

1 physical volume(s) resized or updated / 0 physical volume(s) not resized

# pvs

PV         VG Fmt  Attr PSize    PFree

/dev/sda2  cl lvm2 a–    77.12g   <9.23g

/dev/sdb1     lvm2 —  <150.00g <150.00g

# reboot

 

After reboot there is no more complains and PV size matches /dev/sda2 size.

# vgs

VG #PV #LV #SN Attr   VSize  VFree

cl   1   3   0 wz–n- 77.12g <9.23g

# pvs

PV         VG Fmt  Attr PSize    PFree

/dev/sda2  cl lvm2 a–    77.12g   <9.23g

/dev/sdb1     lvm2 —  <150.00g <150.00g

# fdisk -l /dev/sda

Disk /dev/sda: 78.1 GiB, 83886080000 bytes, 163840000 sectors

Units: sectors of 1 * 512 = 512 bytes

Sector size (logical/physical): 512 bytes / 512 bytes

I/O size (minimum/optimal): 512 bytes / 512 bytes

Disklabel type: dos

Disk identifier: 0x845c40d9

 

Device     Boot   Start       End   Sectors  Size Id Type

/dev/sda1  *       2048   2099199   2097152    1G 83 Linux

/dev/sda2       2099200 165939199 163840000 78.1G 8e Linux LVM

 

 

Anyway it is not true. The Linux box is running in VMware environment and we just need to decrease the size of the virtual disk. Just fire up the webconsole and do it.

There is something wrong:

 

VMware GUI does not want to shrink the volume. Fortunately there are plenty of clever people around and one of them just made following article.

It is for windows, but Linux is easier to do it. The procedure is like this:

  1. Shutdown VM
  2. Login to esxi server as root
  3. Do calculation (80GB = 80 * 1024 * 1024 * 1024 / 512) or take the number directly from fstab 163840000 sectors
  4. Change to directory /vmfs/volumes/<DS-dir>/<VM-dir>
  5. Edit  VM.vmdk file
  6. Change number of sectors like this:

From:

# Extent description

RW 1048576000 VMFS “vm-flat.vmdk”

To:

# Extent description

RW 163840000 VMFS “vm-flat.vmdk”

  1. Migrate VM to a different DS
  2. Start up VM and the size is changed.

# fdisk -l /dev/sda

Disk /dev/sda: 78.1 GiB, 83886080000 bytes, 163840000 sectors

Units: sectors of 1 * 512 = 512 bytes

Sector size (logical/physical): 512 bytes / 512 bytes

I/O size (minimum/optimal): 512 bytes / 512 bytes

Disklabel type: dos

Disk identifier: 0x845c40d9

 

Device     Boot   Start       End   Sectors  Size Id Type

/dev/sda1  *       2048   2099199   2097152    1G 83 Linux

/dev/sda2       2099200 165939199 163840000 78.1G 8e Linux LVM

 

# vgdisplay -v

— Volume group —

VG Name               cl

System ID

Format                lvm2

Metadata Areas        1

Metadata Sequence No  16

VG Access             read/write

VG Status             resizable

MAX LV                0

Cur LV                3

Open LV               3

Max PV                0

Cur PV                1

Act PV                1

VG Size               77.12 GiB

PE Size               4.00 MiB

Total PE              19743

Alloc PE / Size       17381 / 67.89 GiB

Free  PE / Size       2362 / <9.23 GiB

VG UUID               0XLSu8-aOeE-Mdg1-5ViQ-2zaQ-9hPv-X9T4h1

 

— Logical volume —

LV Path                /dev/cl/swap

LV Name                swap

VG Name                cl

LV UUID                3yJjj2-cLdt-uaiT-C3er-i2Zx-u0JY-Ip0xG4

LV Write Access        read/write

LV Creation host, time localhost, 2020-03-12 09:23:04 +0100

LV Status              available

# open                 2

LV Size                7.89 GiB

Current LE             2021

Segments               1

Allocation             inherit

Read ahead sectors     auto

– currently set to     8192

Block device           253:1

 

— Logical volume —

LV Path                /dev/cl/root

LV Name                root

VG Name                cl

LV UUID                LAmWX2-3DSS-VD7d-9rWs-SFzc-hTni-akto3r

LV Write Access        read/write

LV Creation host, time localhost, 2020-03-12 09:23:06 +0100

LV Status              available

# open                 1

LV Size                50.00 GiB

Current LE             12800

Segments               1

Allocation             inherit

Read ahead sectors     auto

– currently set to     8192

Block device           253:0

 

— Logical volume —

LV Path                /dev/cl/home

LV Name                home

VG Name                cl

LV UUID                3NCUe4-JZVY-FdkD-h8mC-Xuuz-LL6M-uslaFo

LV Write Access        read/write

LV Creation host, time host.example.net, 2020-06-16 09:39:03 +0200

LV Status              available

# open                 1

LV Size                10.00 GiB

Current LE             2560

Segments               1

Allocation             inherit

Read ahead sectors     auto

– currently set to     8192

Block device           253:2

 

— Physical volumes —

PV Name               /dev/sda2

PV UUID               fB3mro-9lDl-lMBg-0jNh-qqzx-PReV-iHUu8T

PV Status             allocatable

Total PE / Free PE    19743 / 2362

 

# pvdisplay /dev/sda2 -m

— Physical volume —

PV Name               /dev/sda2

VG Name               cl

PV Size               77.12 GiB / not usable 2.00 MiB

Allocatable           yes

PE Size               4.00 MiB

Total PE              19743

Free PE               2362

Allocated PE          17381

PV UUID               fB3mro-9lDl-lMBg-0jNh-qqzx-PReV-iHUu8T

 

— Physical Segments —

Physical extent 0 to 2020:

Logical volume      /dev/cl/swap

Logical extents     0 to 2020

Physical extent 2021 to 4580:

Logical volume      /dev/cl/home

Logical extents     0 to 2559

Physical extent 4581 to 17380:

Logical volume      /dev/cl/root

Logical extents     0 to 12799

Physical extent 17381 to 19742:

FREE

 

 

Continue reading Shrink CentOS 8.1 XFS disk in VM
Routers

OpenWRT upgrade for TL-WR741ND

Written by mirek Published Discussion 1 Comment on OpenWRT upgrade for TL-WR741ND

I have tried to upgrade my old router TL-WR741ND today. OpenWRT pages report that routers with 4 MB flash memory and 32 MB RAM are not supported anymore(4/32 warning)? It is not recommended to buy them for usage with OpenWRT. Anything with bigger RAM or Flash memory is more stable for new FW’s. Anyway anything newer than

OpenWrt Barrier Breaker 14.07 / LuCI Trunk (0.12+svn-r10530)

is better than the current one. The OpenWRT pages have the latest FW here( 18.06.8).

The upgrade could be done through router’s LuCI GUI after FW image download to the local PC. It is necessary to select the local image file and select Keep settings checkbox to keep the current configuration. The image will be loaded by pressing the Flash Image button:

The load was successful and the router booted up, but LuCI does not work? Fortunately there is a chance to log into the router through ssh (it is a good idea to enable it before FW upgrade? The log reported following errors:

root@OpenWrt:~# logread -l 100
Sun Apr 26 09:31:39 2020 daemon.err uhttpd[905]: /usr/lib/lua/luci/dispatcher.lua:234: /etc/config/luci seems to be corrupt, unable to find section 'main'
Sun Apr 26 09:31:39 2020 daemon.err uhttpd[905]: stack traceback:
Sun Apr 26 09:31:39 2020 daemon.err uhttpd[905]: [C]: in function 'assert'
Sun Apr 26 09:31:39 2020 daemon.err uhttpd[905]: /usr/lib/lua/luci/dispatcher.lua:234: in function 'dispatch'
Sun Apr 26 09:31:39 2020 daemon.err uhttpd[905]: /usr/lib/lua/luci/dispatcher.lua:127: in function </usr/lib/lua/luci/dispatcher.lua:126>
After LuCi config check itis obvious that the error is not related to LuCI configuration?
root@OpenWrt:~# head /etc/config/luci
config core 'main'
option lang 'auto'
option resourcebase '/luci-static/resources'
option mediaurlbase '/luci-static/bootstrap'

The web search unveiled several articles and the combination is going to give me the solution.The first one points out to rpcd error, because LuCI config is fine. A simple test confirms that the rpcd does not work:
root@OpenWrt:~# ps w|grep rpcd
1682 root 1200 S grep rpcd

The next part of the article recommends to use following command, which did not work as well:
root@OpenWrt:~# service rpcd restart
-ash: service: not found

A simple start of rpcd worked well:
root@OpenWrt:~# rpcd &

I have done reboot and LuCI did not work again. The further search found another recommendation – use factory reset via HW button or command:
root@OpenWrt:~# firstboot
This will erase all settings and remove any installed packages. Are you sure? [N/y]
y
/dev/mtdblock3 is mounted as /overlay, only erasing files
root@OpenWrt:~# reboot

It was enough to clean up all configuration and rpcd started automatically after reboot. LuCI works fine now. It is time to buy anew router with more memory?

Continue reading OpenWRT upgrade for TL-WR741ND
Programming, Uncategorized

Uptime and ESP8266

Written by mirek Published Discussion Leave a Comment on Uptime and ESP8266

Any time meassurement in ESP8266 is done through millis() function. A nice article about this topic is here. The article emphasis duration meassuremnt instead of timestamps. The modular arithmetic is going to bypass any rollover of unsigned long type used for millis() function. The rollover happens approximately each 47 days. If we need longer time we can increae it by using 64 bits:

uint64_t millis64() {
    static uint32_t low32, high32;
    uint32_t new_low32 = millis();
    if (new_low32 < low32) high32++;
    low32 = new_low32;
    return (uint64_t) high32 << 32 | low32;
}

If we do not need to measure seconds for uptime we can use unsigned long even for thousand years, by simplification to minutes. It is just necessary to use millis one time per minute (it is 60 000 ticks). The code could look like this: 

unsigned long TT,upM=0,upH=0,upD=0;

if (millis()-TT >60000)
    {
      upM++;
      if (upM == 60) 
      {
        upM=0;
        upH++;
      }
      if (upH == 24)
      {
        upH=0;
        upD++;
      }
    }
Continue reading Uptime and ESP8266
Programming

Operator time for SIM800L

Written by mirek Published

Module SIM800L is simple GSM modem, which could be connected via serial line (even SW serial). It could send/receive SMS and calls. The calls require connected microphone and speaker. SMS sending works fine without additional HW. A nice article (in Czech) is here. The modem needs good quality power input at 4.2V and peak current is 3A.

After power is applied the modem LED is blinking quickly about 10 times. Then it will be off for a while (modem is connecting to GSM network) and after successful connection the speed of LED blinking is slower. SIM should be inserted with contacts pointing to printed board.

Související obrázek

The modem worked fine if there was good power input. The only problem was with time synchronization  to operator’s time. Each sync has ended up with following time:

04/01/01,00:01:55+32

Any parameter change did not help. After some failures I have tried to change CBAND (Get and Set Mobile Operation Band ) parameter. The original setting was:

+CBAND: EGSM_MODE,ALL_BAND

The modem reported following options for CBAND parameter:

+CBAND: (EGSM_MODE,DCS_MODE,GSM850_MODE,PCS_MODE,EGSM_DCS_MODE, GSM850_PCS_MODE,EGSM_PCS_MODE,ALL_BAND)

I have tried “DCS_MODE” and it did not work, but the second try “EGSM_MODE” or “ALL_BAND” worked fine.

Change to EGSM_MODE:

at+cband=EGSM_MODE

OK

*PSUTTZ: 2018,10,31,17,15,19,”+4″,0

DST: 0

+CIEV: 10,”23002″,”CEZ”,”CEZ”, 0, 0

Change check:

at+cband?

+CBAND: EGSM_MODE

OK

and operator time is?

at+cclk?

+CCLK: “18/10/31,18:15:32+04”

After storing new settings into modem config (AT&W) the modem picks up the operator time immediately after power connection and connecting to GSM network.

Continue reading Operator time for SIM800L
web

Howto add animated gif to theme Twenty Sixteen Header Image

Written by mirek Published Discussion Leave a Comment on Howto add animated gif to theme Twenty Sixteen Header Image

WordPress theme Twenty Sixteen is not able to upload animated gif as header picture. It is always going to change the picture to resized gif. This gif has only one frame and it is without a single movement.

It does not matter if you select Crop image or Skip cropping it is always cropped to following sizes 1200×280, 1024×238, 768×180, 300×70, 150×150. These pictures are more jpg than gif:-)

The easiest way how to change it is to find the location of the affected picture via right click and selecting “Copy image address”. It is going to give you a link on your website something like this:

https://www.example.ccom/wp-content/uploads/2018/10/pict.gif

if the original name was pict.gif. If you check the location on your web server you will find files with the different size for the same picture. They will be named gif, but there will be no animation. Just use resize service:

https://ezgif.com/resize

and make proper sizes of your original animated gif. Then upload them to the same directory on your web server as it points by link above. The replacement of static gif’s by the animated will work fine and you have animated gif in your header.

The normal way how to include animated gif’s in the posts works fine and the steps above are necessary only for the header file;-)

Continue reading Howto add animated gif to theme Twenty Sixteen Header Image
phone