Días de Linux: 09/01/2023

Actualizar de Kamikaze 8 a backfire 10.03.1

Conecto SSH con el router WRT54GL con el comando:

ssh -oKexAlgorithms=+diffie-hellman-group1-sha1 192.168.110.1

Al hacer

opkg update

aparece una lista de todos los paquetes que no puede ser actualizados porque corresponden a una arquitectura diferente.

edito el archivo /etc/opkg.conf

vi /etc/opkg.conf

que contiene:

src/gz snapshots http://downloads.openwrt.org/kamikaze/8.09.2/brcm-2.4/packages
dest root /
dest ram /tmp
lists_dir ext /var/opkg-lists
option overlay_root /jffs

actualizo la primera linea y la sustituyo por:mount

src/gz snapshots http://downloads.openwrt.org/backfire/10.03.1/brcm-2.4/packages

y ejecuto dos veces:

opkg update

Despues al intentar instalar openvpn desde el UI LuCI aparece el siguiente error:

Installing openvpn (2.1.4-3) to root...

Collected errors:
* verify_pkg_installable: Only have 0kb available on filesystem /, pkg openvpn needs 185
* opkg_install_cmd: Cannot install package openvpn.

root@GW8923:~# df -h
Filesystem Size Used Available Use% Mounted on
/dev/root 1.7M 1.7M 0 100% /rom
tmpfs 7.0M 1.2M 5.8M 18% /tmp
/dev/mtdblock/4 1.5M 336.0K 1.2M 22% /overlay
mini_fo:/overlay 1.7M 1.7M 0 100% /

root@GW8923:~# cp /rom/etc/opkg.conf /overlay/etc/opkg.conf

root@GW8923:~# df -h
Filesystem Size Used Available Use% Mounted on
/dev/root 1.7M 1.7M 0 100% /rom
tmpfs 7.0M 1.3M 5.7M 18% /tmp
/dev/mtdblock/4 1.5M 336.0K 1.2M 22% /overlay
mini_fo:/overlay 1.7M 1.7M 0 100% /

renombro overlay/etc a /overlay/etc.old como paso previo para eliminar ese directorio

root@GW8923:~# mv /overlay/etc/ /overlay/etc.old

sigue apareciendo que hay 0 espacio disponible:

root@GW8923:~# df -h
Filesystem Size Used Available Use% Mounted on
/dev/root 1.7M 1.7M 0 100% /rom
tmpfs 7.0M 1.4M 5.6M 20% /tmp
/dev/mtdblock/4 1.5M 336.0K 1.2M 22% /overlay
mini_fo:/overlay 1.7M 1.7M 0 100% /

Sin embargo, intento instalar de nuevo el paquete openvpn

Installing openvpn (2.1.4-3) to root...
Downloading http://downloads.openwrt.org/backfire/10.03.1/brcm-2.4/packages/openvpn_2.1.4-3_brcm-2.4.ipk.
Installing kmod-tun (2.4.37.9-1) to root...
Downloading http://downloads.openwrt.org/backfire/10.03.1/brcm-2.4/packages/kmod-tun_2.4.37.9-1_brcm-2.4.ipk.
Installing libopenssl (0.9.8r-1) to root...
Downloading http://downloads.openwrt.org/backfire/10.03.1/brcm-2.4/packages/libopenssl_0.9.8r-1_brcm-2.4.ipk.
Installing zlib (1.2.3-5) to root...
Downloading http://downloads.openwrt.org/backfire/10.03.1/brcm-2.4/packages/zlib_1.2.3-5_brcm-2.4.ipk.
Installing liblzo (2.04-1) to root...
Downloading http://downloads.openwrt.org/backfire/10.03.1/brcm-2.4/packages/liblzo_2.04-1_brcm-2.4.ipk.
Configuring kmod-tun.
Configuring zlib.
Configuring libopenssl.
Configuring liblzo.
Configuring openvpn.

Collected errors:
* resolve_conffiles: Existing conffile /etc/config/openvpn is different from the conffile in the new package. The new conffile will be placed at /etc/config/openvpn-opkg.

Despues de instalarlo el espacio disponible en /overlay es menor:

root@GW8923:~# df -h
Filesystem Size Used Available Use% Mounted on
/dev/root 1.7M 1.7M 0 100% /rom
tmpfs 7.0M 1.3M 5.7M 19% /tmp
/dev/mtdblock/4 1.5M 1.3M 240.0K 84% /overlay
mini_fo:/overlay 1.7M 1.7M 0 100% /

Instalo luci-app-openvpn

y el espacio disponible queda en:

root@GW8923:~# df -h

Filesystem Size Used Available Use% Mounted on

/dev/root 1.7M 1.7M 0 100% /rom

tmpfs 7.0M 1.3M 5.7M 19% /tmp

/dev/mtdblock/4 1.5M 1.3M 180.0K 88% /overlay

mini_fo:/overlay 1.7M 1.7M 0 100% /

el sistema de archivos tiene los siguientes puntos de montaje:

  
root@GW8923:~# mount

rootfs on / type rootfs (rw)

/dev/root on /rom type squashfs (ro)

none on /dev type devfs (rw)

none on /proc type proc (rw)

tmpfs on /tmp type tmpfs (rw,nosuid,nodev)

devpts on /dev/pts type devpts (rw)

/dev/mtdblock/4 on /overlay type jffs2 (rw)

mini_fo:/overlay on / type mini_fo (rw)

Mi actual configuración de almacenamiento en Fedora 29 tiene dos RAID0. Uno con discos SATA y bcache y el segundo con discos SDD. El RAID0 esta implementado con Intel Rapid Storage Technology (RST).

Para mostrar el listado de unidades de almacenamiento y particiones actual, ejecutar el comando:

lsblk -o NAME,MAJ:MIN,RM,SIZE,TYPE,FSTYPE,MOUNTPOINT

NAME          MAJ:MIN RM   SIZE TYPE  FSTYPE          MOUNTPOINT
sda             8:0    0 232.9G disk  isw_raid_member 
└─md126         9:126  0 465.8G raid0                 
  ├─md126p1   259:0    0   120G md    xfs             /
  ├─md126p2   259:1    0   1.9G md    ext4            
  └─md126p3   259:2    0 343.9G md    xfs             /home/VMsRepository
sdb             8:16   0 232.9G disk  isw_raid_member 
└─md126         9:126  0 465.8G raid0                 
  ├─md126p1   259:0    0   120G md    xfs             /
  ├─md126p2   259:1    0   1.9G md    ext4            
  └─md126p3   259:2    0 343.9G md    xfs             /home/VMsRepository
sdc             8:32   0 465.8G disk  isw_raid_member 
└─md124         9:124  0 931.5G raid0                 
  └─md124p1   259:3    0 931.5G md    bcache          
    └─bcache0 252:0    0 931.5G disk  xfs             /home
sdd             8:48   0 111.8G disk                  
├─sdd1          8:49   0   500M part  vfat            /boot/efi
├─sdd2          8:50   0     1G part  ext4            /boot
├─sdd3          8:51   0   7.9G part  swap            [SWAP]
└─sdd4          8:52   0 102.5G part  bcache          
  └─bcache0   252:0    0 931.5G disk  xfs             /home
sde             8:64   0 465.8G disk  isw_raid_member 
└─md124         9:124  0 931.5G raid0                 
  └─md124p1   259:3    0 931.5G md    bcache          
    └─bcache0 252:0    0 931.5G disk  xfs             /home
sdk             8:160  0 931.5G disk                  
└─sdk1          8:161  0 931.5G part  ext4            /run/media/Backup01
sr0            11:0    1  1024M rom

Configuración RAID0 con discos SATA:

# mdadm --detail /dev/md124
/dev/md124:
         Container : /dev/md/imsm0, member 0
        Raid Level : raid0
        Array Size : 976768000 (931.52 GiB 1000.21 GB)
      Raid Devices : 2
     Total Devices : 2

             State : clean 
    Active Devices : 2
   Working Devices : 2
    Failed Devices : 0
     Spare Devices : 0

        Chunk Size : 128K

Consistency Policy : none


              UUID : 29a34795:f040ca44:6d80b65e:22c429a8
    Number   Major   Minor   RaidDevice State
       1       8       64        0      active sync   /dev/sde
       0       8       32        1      active sync   /dev/sdc

# mdadm --detail /dev/md125
/dev/md125:
           Version : imsm
        Raid Level : container
     Total Devices : 2

   Working Devices : 2


              UUID : c0bdc8e8:5bba49c4:0467a19c:d0a09f7a
     Member Arrays : /dev/md/Volume0_0

    Number   Major   Minor   RaidDevice

       -       8       64        -        /dev/sde
       -       8       32        -        /dev/sdc

Y la configuración de RAID0 con discos SSD:

# mdadm --detail /dev/md126
/dev/md126:
         Container : /dev/md/imsm1, member 0
        Raid Level : raid0
        Array Size : 488391680 (465.77 GiB 500.11 GB)
      Raid Devices : 2
     Total Devices : 2

             State : clean 
    Active Devices : 2
   Working Devices : 2
    Failed Devices : 0
     Spare Devices : 0

        Chunk Size : 128K

Consistency Policy : none


              UUID : dc00ef36:c7b058c3:a6598f75:06f6e016
    Number   Major   Minor   RaidDevice State
       1       8        0        0      active sync   /dev/sda
       0       8       16        1      active sync   /dev/sdb

# mdadm --detail /dev/md127
/dev/md127:
           Version : imsm
        Raid Level : container
     Total Devices : 2

   Working Devices : 2


              UUID : 0125bb79:4caae511:15023fbb:4006b23c
     Member Arrays : /dev/md/SystemRAID0

    Number   Major   Minor   RaidDevice

       -       8        0        -        /dev/sda
       -       8       16        -        /dev/sdb

Comparando los dos RAID0:
Primero en el raid SSD

#sudo fio --name=randread --ioengine=libaio --iodepth=16 --rw=randread --bs=4k --direct=0 --size=512M --numjobs=4 --runtime=240 --group_reporting

randread: (g=0): rw=randread, bs=(R) 4096B-4096B, (W) 4096B-4096B, (T) 4096B-4096B, ioengine=libaio, iodepth=16
...
fio-3.7
Starting 4 processes
randread: Laying out IO file (1 file / 512MiB)
randread: Laying out IO file (1 file / 512MiB)
randread: Laying out IO file (1 file / 512MiB)
randread: Laying out IO file (1 file / 512MiB)
Jobs: 4 (f=4): [r(4)][100.0%][r=153MiB/s,w=0KiB/s][r=39.1k,w=0 IOPS][eta 00m:00s]
randread: (groupid=0, jobs=4): err= 0: pid=19959: Sun May 26 20:38:56 2019
   read: IOPS=40.6k, BW=159MiB/s (166MB/s)(2048MiB/12902msec)
    slat (usec): min=73, max=8454, avg=94.76, stdev=28.83
    clat (usec): min=2, max=10144, avg=1477.35, stdev=157.81
     lat (usec): min=87, max=10244, avg=1572.52, stdev=164.40
    clat percentiles (usec):
     |  1.00th=[ 1270],  5.00th=[ 1303], 10.00th=[ 1336], 20.00th=[ 1385],
     | 30.00th=[ 1401], 40.00th=[ 1434], 50.00th=[ 1467], 60.00th=[ 1483],
     | 70.00th=[ 1516], 80.00th=[ 1565], 90.00th=[ 1631], 95.00th=[ 1696],
     | 99.00th=[ 1860], 99.50th=[ 1958], 99.90th=[ 2278], 99.95th=[ 2966],
     | 99.99th=[ 9765]
   bw (  KiB/s): min=38320, max=42080, per=25.03%, avg=40680.80, stdev=871.14, samples=100
   iops        : min= 9580, max=10520, avg=10170.18, stdev=217.76, samples=100
  lat (usec)   : 4=0.01%, 10=0.01%, 20=0.01%, 100=0.01%, 250=0.01%
  lat (usec)   : 500=0.01%, 750=0.01%, 1000=0.01%
  lat (msec)   : 2=99.59%, 4=0.38%, 10=0.01%, 20=0.01%
  cpu          : usr=4.53%, sys=14.80%, ctx=524767, majf=0, minf=111
  IO depths    : 1=0.1%, 2=0.1%, 4=0.1%, 8=0.1%, 16=100.0%, 32=0.0%, >=64=0.0%
     submit    : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
     complete  : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.1%, 32=0.0%, 64=0.0%, >=64=0.0%
     issued rwts: total=524288,0,0,0 short=0,0,0,0 dropped=0,0,0,0
     latency   : target=0, window=0, percentile=100.00%, depth=16

Run status group 0 (all jobs):
   READ: bw=159MiB/s (166MB/s), 159MiB/s-159MiB/s (166MB/s-166MB/s), io=2048MiB (2147MB), run=12902-12902msec

Disk stats (read/write):
    md126: ios=520885/0, merge=0/0, ticks=0/0, in_queue=0, util=0.00%, aggrios=262144/0, aggrmerge=0/0, aggrticks=21902/0, aggrin_queue=27, aggrutil=99.02%
  sdb: ios=262144/0, merge=0/0, ticks=22610/0, in_queue=44, util=99.02%
  sda: ios=262144/0, merge=0/0, ticks=21195/0, in_queue=10, util=97.82%

Ahora en el raid SATA con bcache

# sudo fio --name=randread --ioengine=libaio --iodepth=16 --rw=randread --bs=4k --direct=0 --size=512M --numjobs=4 --runtime=240 --group_reporting

randread: (g=0): rw=randread, bs=(R) 4096B-4096B, (W) 4096B-4096B, (T) 4096B-4096B, ioengine=libaio, iodepth=16
...
fio-3.7
Starting 4 processes
randread: Laying out IO file (1 file / 512MiB)
randread: Laying out IO file (1 file / 512MiB)
randread: Laying out IO file (1 file / 512MiB)
randread: Laying out IO file (1 file / 512MiB)
Jobs: 4 (f=4): [r(4)][100.0%][r=1164KiB/s,w=0KiB/s][r=291,w=0 IOPS][eta 00m:00s]
randread: (groupid=0, jobs=4): err= 0: pid=20549: Sun May 26 20:44:54 2019
   read: IOPS=291, BW=1168KiB/s (1196kB/s)(274MiB/240041msec)
    slat (usec): min=80, max=194852, avg=13690.49, stdev=10673.08
    clat (usec): min=5, max=610985, avg=205391.13, stdev=44617.88
     lat (msec): min=27, max=635, avg=219.08, stdev=46.30
    clat percentiles (msec):
     |  1.00th=[  124],  5.00th=[  142], 10.00th=[  155], 20.00th=[  169],
     | 30.00th=[  180], 40.00th=[  190], 50.00th=[  201], 60.00th=[  211],
     | 70.00th=[  224], 80.00th=[  241], 90.00th=[  264], 95.00th=[  284],
     | 99.00th=[  338], 99.50th=[  355], 99.90th=[  401], 99.95th=[  430],
     | 99.99th=[  575]
   bw (  KiB/s): min=   80, max=  448, per=24.99%, avg=291.67, stdev=44.48, samples=1920
   iops        : min=   20, max=  112, avg=72.88, stdev=11.12, samples=1920
  lat (usec)   : 10=0.01%
  lat (msec)   : 50=0.01%, 100=0.07%, 250=85.32%, 500=14.56%, 750=0.02%
  cpu          : usr=0.08%, sys=0.39%, ctx=70109, majf=0, minf=109
  IO depths    : 1=0.1%, 2=0.1%, 4=0.1%, 8=0.1%, 16=99.9%, 32=0.0%, >=64=0.0%
     submit    : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
     complete  : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.1%, 32=0.0%, 64=0.0%, >=64=0.0%
     issued rwts: total=70076,0,0,0 short=0,0,0,0 dropped=0,0,0,0
     latency   : target=0, window=0, percentile=100.00%, depth=16

Run status group 0 (all jobs):
   READ: bw=1168KiB/s (1196kB/s), 1168KiB/s-1168KiB/s (1196kB/s-1196kB/s), io=274MiB (287MB), run=240041-240041msec

Disk stats (read/write):
    bcache0: ios=70048/212, merge=0/0, ticks=954259/668, in_queue=954927, util=28.96%, aggrios=35128/266, aggrmerge=0/6, aggrticks=206/200, aggrin_queue=303, aggrutil=0.21%
    md124: ios=70028/331, merge=0/0, ticks=0/0, in_queue=0, util=0.00%, aggrios=35014/155, aggrmerge=0/14, aggrticks=477297/315, aggrin_queue=460121, aggrutil=21.90%
  sde: ios=33713/152, merge=0/13, ticks=448824/272, in_queue=432278, util=20.81%
  sdc: ios=36316/158, merge=0/15, ticks=505771/359, in_queue=487965, util=21.90%
  sdd: ios=229/202, merge=0/13, ticks=412/401, in_queue=606, util=0.21%

Usando la utilidad de benchmarking de gnome-disks.

En la partición 3 del raid SSD:

En la partición con bcache del raid con discos SATA:

Primero con la cache vacía:

Después de realizar el test anterior, la cache ya contiene datos y el tiempo de acceso se reduce de forma drástica:

Comparamos ahora el mismo dispositivo sin cache activada:

Con respecto al tiempo de arranque:
Con la configuración actual de discos HDD en RAID0 con bcache en SSD, el tiempo de arranque es:

# systemd-analyze 
Startup finished in 1.824s (kernel) + 1.994s (initrd) + 58.781s (userspace) = 1min 2.600s
graphical.target reached after 58.764s in userspace

Un segundo arranque confirma resultados similares:

# systemd-analyze 
Startup finished in 1.793s (kernel) + 1.874s (initrd) + 59.652s (userspace) = 1min 3.320s
graphical.target reached after 59.584s in userspace

Para obtener el resultado de forma gráfica, se puede usar:

# systemd-analyze plot > Pictures/Bootcharts/plot-20190603-1.svg

o también:

/usr/lib/systemd/systemd-bootchart

que crea un archivo svg en /run/log

En este articulo se citan varias posibilidades para reducir el tiempo de arranque.

Días de Linux

viernes, 1 de septiembre de 2023

Actualizando el firmware de OpenWRT en el router WRT54GL

Rendimiento de discos SATA con bcache y SSD