tuxad blog 2022-10-16T22:46:46+02:00 Frank W. Bergmann /blog /blog/ NanoBlogger Was ist Digitalisierung? Frank W. Bergmann /blog/archives/2022/10/16/was_ist_digitalisierung/index.html 2022-10-16T13:52:24+02:00 2022-10-16T13:52:24+02:00

Die Video-AG des Engelsburg-Gymnasiums hatte der hessischen "Digitalministerin" Prof. Dr. Kristina Sinemus ein paar Fragen gestellt. Unter anderem stellten die Kinder die Frage: "Was ist Digitalisierung eigentlich?"

Das war ihre Antwort:

Digitalisierung ist eine Möglichkeit, um über Infrastruktur, also die Anwendung sozusagen auf den Glasfasernetzen, neue Ideen in die Umsetzung zu bringen. Digitalisierung heißt Anschlüsse und mathematische Algorithmen - die Einsen und die Nullen.

Alternativ dazu die spontane Antwort einer 15-jährigen Schülerin auf dieselbe Frage:

Wenn man zum Beispiel tägliche Handlungen vereinfacht, indem man sie von analog zu digital umwandelt.

Diese spontane Antwort der Schülerin entspricht ungefähr einer vereinfachten Definition auf digital-magazin.de:

Die einfachste Definition von Digitalisierung besagt, dass analoge Inhalte oder Prozesse in eine digitale Form oder Arbeitsweise umgewandelt werden.

Die Erklärung von Frau Prof. Dr. Sinemus ist nicht nur leicht am Thema vorbei, sondern sogar falsch. Digitalisierung ist unabhängig von Netzen und erst recht sind Glasfasernetze keine Voraussetzung für Digitalisierung. Wenn jemand z.B. Anfang der 80er Jahre auf einem Apple II (=unvernetzt) ein Textdokument erstellt hat, das vorher nur handschriftlich vorlag, und das dann mehrfach ausgedruckt hat, dann stellte das bereits eine "Digitalisierung" dar. Ebenso werden nicht zwingend Algorithmen benötigt für Verarbeitung, Speicherung und Übertragung digitaler Daten. Und "Anschlüsse" sind ebenso irrelevant.

Referenz: 4 Fragen an die Hessische Digitalministerin

Jitsi Meet RPM packages for RHEL/CentOS/OEL 7 Frank W. Bergmann /blog/archives/2021/01/24/jitsi_meet_rpm_packages_for_rhelcentosoel_7/index.html 2021-01-24T18:19:18+02:00 2021-01-24T18:19:18+02:00

First enable tuxad repository: 


rpm -i http://www.tuxad.com/repo/7/tuxad.rpm

We also need the EPEL repository. If you don't have it configured then follow the instructions on fedoraproject.org/wiki/EPEL.

Next required repo is nodesource. Enable it via tuxad meta package: 


yum install nodesource-release-el7

Install jitsi-meet-install package which will by dependency also install other required packages: 


yum install jitsi-meet-install

Add two entries to your DNS server to point to the server you are installing on:

  • jitsi.YOURDOMAIN.COM
  • auth.jitsi.YOURDOMAIN.COM

After DNS is set up the final Jitsi Meet configuration can be done. The configuring script will call certbot to get certificates from Let's Encrypt.

It is important that (at least when running the configure script) no service is running on TCP port 80. If you have an http server running on this port then shut it down, run the configure script and then start it up again.

The configure script requires at least the IP address (IPv4) and the hostname. Example: 


jitsi-meet-install.sh 1.2.3.4 jitsi.example.com

Run this script with your own settings. You are done!

Power Measuring several SSDs (Crucial, Kingston, Apacer) Frank W. Bergmann /blog/archives/2020/03/23/power_measuring_several_ssds_crucial_kingston_apacer/index.html 2020-03-23T01:48:49+02:00 2020-03-23T01:48:49+02:00

I was curious if the widely used Crucial BX 500 did also "fail" as 480 GB model. And actually it shows the same picture as the 240 GB model: Fine consumption on standby, read values "ok" but high on write - with the same high bandwidth from approx. 1 Watts to 3.3 Watts.

Another two Patriot Burst 240 did also show the same results as the first Patriot Burst 240 measurement.

And the Kingston A400 960 was the first "1 TB class" SSD I measured. It did show the same great values like the 480 model did.

In addition to even more SSDs I made also the first measurement of an M.2 SSD (SATA, of course). Values were the same as for 2.5" SSD.

Apacer AS340 240 GB

240 GB means actually 468862128 sectors.

  • Idle 0.6 W
  • Read 0.76 W
  • Write 1.25 W, max 3.3 W
  • Read all sectors: 10m15.074s (372 MiB/s)
  • Write all sectors: 50m43.854s (75 MiB/s)

x-y diagram of measurements

Crucial BX500 480

480 GB means actually 937703088 sectors.

  • Idle 0.49 W
  • Read 0.86 W
  • Write 1.66 W
  • Read all sectors: 20m25.698s (373 MiB/s)
  • Write all sectors: 94m0.830s (81 MiB/s)

x-y diagram of measurements

Kingston A400 960

960 GB means actually 1875385008 sectors.

  • Idle 0.28 W
  • Read 0.49 W
  • Write 0.49 W
  • Read all sectors: 37m22.774s (408 MiB/s)
  • Write all sectors: 44m10.833s (345 MiB/s)

x-y diagram of measurements

M.2 Kingston A400 480

480 GB means actually 937703088 sectors.

  • Idle 0.3 W
  • Read 0.5 W
  • Write 0.5 W
  • Read all sectors: 20m29.645s (372 MiB/s)
  • Write all sectors: 23m11.808s (329 MiB/s)

x-y diagram of measurements

Values overview

Kingston A400 960
SATA SSD r pow
(W)		 w pow
(W)		 r spd
(MB/s)		 w spd
(MB/s)
ADATA SU800 256 1.4 2.2
Apacer AS340 240 0.76 1.25 372 75
Corsair Force LE 120 0.58 0.58
Crucial BX500 240 0.6 1.7 372 99
Crucial BX500 480 0.86 1.66 373 81
Kingston A400 240 0.52 1.08 371 290
Kingston A400 480 0.5 0.5 371 328
0.49 0.49 408 345
Patriot Burst 120 0.54 0.54 371 86
Patriot Burst 240 0.52 0.52 369 327
Patriot Burst 480 0.51 0.5 371 328
Samsung MZ7LN128 0.97 1.57 353 150
Sandisk Z410 120 0.7 1.5
Verbatim Vi500 S3 240 0.59 0.98 211 117
Verbatim Vi500 S3 480 0.84 1.13 359 129
WD Green 120 0.96 1.3
M.2 SATA r pow
(W)		 w pow
(W)		 r spd
(MB/s)		 w spd
(MB/s)
Kingston A400 480 0.5 0.5 372 329
2.5 HD r pow
(W)		 w pow
(W)		 r spd
(MB/s)		 w spd
(MB/s)
Seagate ST1000LM048 2.7 2.6
Toshiba MQ02ABD100H 2.6 3.0
WD WD10SPZX 1.62 2.47 256
3.5 HD r pow
(W)		 w pow
(W)		 r spd
(MB/s)		 w spd
(MB/s)
Seagate ST1000DM010 2.69 2.29
Toshiba DT01ACA100 2.86 2.94

Method

Measuring environment is explained in power measuring ADATA SU800. The calculation of speed as logged in latest measurements was simply done by taking the time for reading and writing the whole SSD in relation to its size.

Other measurements

 Power Measuring Crucial BX500 240 Frank W. Bergmann /blog/archives/2020/02/02/power_measuring_crucial_bx500_240/index.html 2020-02-02T14:31:58+02:00 2020-02-02T14:31:58+02:00

I ordered a Crucial BX250 because it was several times recommended by german magazine c't (Heise Verlag). It was also part of a recommendation for a DYI low power system "7-Watt-Mini" in c't 24/2019.

Unfortunately I can't understand this recommendation: The SSD has low write transfer rates and has a high power consumption on write. And even its (low) power consumption on read is higher than Corsair, Kingston and Patriot have.

Crucial BX500 240

240 GB means actually 468862128 sectors.

  • Idle 0.37 W
  • Read 0.6 W
  • Write 1.74 W, max 3.3 W
  • Read all sectors: 10m14.784s (372 MiB/s)
  • Write all sectors: 38m32.850s (99 MiB/s)

x-y diagram of measurements

Values overview

SATA SSD r pow
(W)		 w pow
(W)		 r spd
(MB/s)		 w spd
(MB/s)
ADATA SU800 256 1.4 2.2
Corsair Force LE 120 0.58 0.58
Crucial BX500 240 0.6 1.7 372 99
Kingston A400 480 0.5 0.5 371 328
Kingston A400 240 0.52 1.08 371 290
Patriot Burst 120 0.54 0.54 371 86
Patriot Burst 240 0.52 0.52 369 327
Patriot Burst 480 0.51 0.5 371 328
Samsung MZ7LN128 0.97 1.57 353 150
Sandisk Z410 120 0.7 1.5
Verbatim Vi500 S3 240 0.59 0.98 211 117
Verbatim Vi500 S3 480 0.84 1.13 359 129
WD Green 120 0.96 1.3
2.5 HD r pow
(W)		 w pow
(W)		 r spd
(MB/s)		 w spd
(MB/s)
Seagate ST1000LM048 2.7 2.6
Toshiba MQ02ABD100H 2.6 3.0
WD WD10SPZX 1.62 2.47 256
3.5 HD r pow
(W)		 w pow
(W)		 r spd
(MB/s)		 w spd
(MB/s)
Seagate ST1000DM010 2.69 2.29
Toshiba DT01ACA100 2.86 2.94

Method

Measuring environment is explained in power measuring ADATA SU800. The calculation of speed as logged in latest measurements was simply done by taking the time for reading and writing the whole SSD in relation to its size.

Other measurements

 Power Measuring Kingston A400 480 and Samsung MZ7LN128 Frank W. Bergmann /blog/archives/2020/01/25/power_measuring_kingston_a400_480_and_samsung_mz7ln128/index.html 2020-01-25T16:00:49+02:00 2020-01-25T16:00:49+02:00

In this session I measured a Kingston A400 480 which is much better than the smaller sizes. It's as great as the Patriot Burst 480: Very fast and very low power consumption. Additionally an old (used) Samsung 128 was measured.

Kingston A400 480

480 GB means actually 937703088 sectors.

  • Idle 0.3 W
  • Read 0.5 W
  • Write 0.5 W
  • Read all sectors: 20m31.041s (371 MiB/s)
  • Write all sectors: 23m12.412s (328 MiB/s)

x-y diagram of measurements

Samsung MZ7LN128

128 GB means actually 250069680 sectors.

  • Idle 0.28 W
  • Read 0.97 W, max 1.59 W
  • Write 1.57 W
  • Read all sectors: 5m45.429s (353 MiB/s)
  • Write all sectors: 13m32.761s (150 MiB/s)

x-y diagram of measurements

Values overview

SATA SSD r pow
(W)		 w pow
(W)		 r spd
(MB/s)		 w spd
(MB/s)
ADATA SU800 256 1.4 2.2
Corsair Force LE 120 0.58 0.58
Kingston A400 120 0.51 1.03 371
Kingston A400 480 0.5 0.5 371 328
Kingston A400 240 0.52 1.08 371 290
Patriot Burst 120 0.54 0.54 371 86
Patriot Burst 240 0.52 0.52 369 327
Patriot Burst 480 0.51 0.5 371 328
Samsung MZ7LN128 0.97 1.57 353 150
Sandisk Z410 120 0.7 1.5
Verbatim Vi500 S3 240 0.59 0.98 211 117
Verbatim Vi500 S3 480 0.84 1.13 359 129
WD Green 120 0.96 1.3
2.5 HD r pow
(W)		 w pow
(W)		 r spd
(MB/s)		 w spd
(MB/s)
Seagate ST1000LM048 2.7 2.6
Toshiba MQ02ABD100H 2.6 3.0
WD WD10SPZX 1.62 2.47 256
3.5 HD r pow
(W)		 w pow
(W)		 r spd
(MB/s)		 w spd
(MB/s)
Seagate ST1000DM010 2.69 2.29
Toshiba DT01ACA100 2.86 2.94

Method

Measuring environment is explained in power measuring ADATA SU800. The speed as logged and calculated in latest measurements was simply done by taking the time for reading and writing the whole SSD in relation to its size.

Other measurements

 Power Measuring Patriot Burst 480 and WD WD10SPZX Frank W. Bergmann /blog/archives/2020/01/04/power_measuring_patriot_burst_480_and_wd_wd10spzx/index.html 2020-01-04T12:23:20+02:00 2020-01-04T12:23:20+02:00

Finally the third Patriot Burst was tested. The results for the 480 are the same like for the 240: Lowest power consumption and highest speed. Additionally a 2.5 hard disk from WD was tested and two 3.5 disks to see the difference between 2.5 and 3.5.

Patriot Burst 480

480 GB means actually 937703088 sectors.

  • Idle 0.28 W
  • Read 0.51 W
  • Write 0.5 W
  • Read all sectors: 20m32.578s (371 MiB/s)
  • Write all sectors: 23m15.145s (328 MiB/s)

x-y diagram of measurements

WD WD10SPZX

This is a 2.5 hard disk with 1953525168 sectors.

  • Idle 0.6 W
  • Read 1.62 W
  • Write 2.47 W
  • Read all sectors: 62m5.537s (256 MiB/s)

x-y diagram of measurements

Toshiba DT01ACA100

This is a 3.5 hard disk with 1 TB.

  • Idle 1.15 W
  • Read 2.86 W
  • Write 2.94 W

x-y diagram of measurements

Seagate ST1000DM010

This is a 3.5 hard disk with 1 TB.

  • Idle 1.26 W
  • Read 2.69 W
  • Write 2.29 W

x-y diagram of measurements

Values overview

SATA SSD r pow
(W)		 w pow
(W)		 r spd
(MB/s)		 w spd
(MB/s)
ADATA SU800 256 1.4 2.2
Corsair Force LE 120 0.58 0.58
Kingston A400 120 0.51 1.03 371
Kingston A400 240 0.52 1.08 371 290
Patriot Burst 120 0.54 0.54 371 86
Patriot Burst 240 0.52 0.52 369 327
Patriot Burst 480 0.51 0.5 371 328
Sandisk Z410 120 0.7 1.5
Verbatim Vi500 S3 240 0.59 0.98 211 117
Verbatim Vi500 S3 480 0.84 1.13 359 129
WD Green 120 0.96 1.3
2.5 HD r pow
(W)		 w pow
(W)		 r spd
(MB/s)		 w spd
(MB/s)
Seagate ST1000LM048 2.7 2.6
Toshiba MQ02ABD100H 2.6 3.0
WD WD10SPZX 1.62 2.47 256
3.5 HD r pow
(W)		 w pow
(W)		 r spd
(MB/s)		 w spd
(MB/s)
Seagate ST1000DM010 2.69 2.29
Toshiba DT01ACA100 2.86 2.94

Method

Measuring environment is explained in power measuring ADATA SU800. The speed as logged and calculated in latest measurements was simply done by taking the time for reading and writing the whole SSD in relation to its size.

Other measurements

 Power Measuring Patriot Burst 240 and Verbatim Vi500 S3 240/480 Frank W. Bergmann /blog/archives/2019/12/16/power_measuring_patriot_burst_240_and_verbatim_vi500_s3_240480/index.html 2019-12-16T22:10:19+02:00 2019-12-16T22:10:19+02:00

This time three SSDs were measured. The Patriot Burst 240 is an exceptional SSD of all measured ones so far: It has the lowest power consumption combined with the highest speed - and it was also the cheapest one.

Patriot Burst 240

240 GB means actually 468862128 sectors.

  • Idle 0.3 W
  • Read 0.52 W
  • Write 0.5 - 0.53 W, avg. 0.52 W
  • Read all sectors: 10m19.060s (369 MB/s)
  • Write all sectors: 11m39.184s (327 MB/s)

x-y diagram of measurements

Verbatim Vi500 S3 240 GB

  • Idle 0.38 W
  • Read 0.59 W
  • Write 0.77 - 1.15 W, avg. 0.98 W
  • Read all sectors: 17m59.678s (211 MB/s)
  • Write all sectors: 32m23.151s (117 MB/s)

x-y diagram of measurements

Verbatim Vi500 S3 480 GB

480 GB means actually 937703088 sectors.

  • Idle 0.38 W
  • Read 0.84 W
  • Write 0.72 - 1.65 W, avg. 1.13 W
  • Read all sectors: 21m14.385s (359 MB/s)
  • Write all sectors: 59m9.244s (129 MB/s)

x-y diagram of measurements

Values overview

read write r speed w speed
Sandisk Z410 120 0.7 W 1.5 W
ADATA SU800 256 1.4 W 2.2 W
WD Green 120 0.96 W 1.3 W
Kingston A400 120 0.51 W 1.03 W 371 MB/s
Corsair Force LE 120 0.58 W 0.58 W
Patriot Burst 120 0.54 W 0.54 W 371 MB/s 86 MB/s
Kingston A400 240 0.52 W 1.08 W 371 MB/s 290 MB/s
Patriot Burst 240 0.52 W 0.52 W 369 MB/s 327 MB/s
Verbatim Vi500 S3 240 0.59 W 0.98 W 211 MB/s 117 MB/s
Verbatim Vi500 S3 480 0.84 W 1.13 W 359 MB/s 129 MB/s

Method

Measuring environment is explained in power measuring ADATA SU800. The speed as logged and calculated in latest measurements was simply done by taking the time for reading and writing the whole SSD in relation to its size.

Other measurements

 ODROID GO as network tool Frank W. Bergmann /blog/archives/2019/09/22/odroid_go_as_network_tool/index.html 2019-09-22T22:47:24+02:00 2019-09-22T22:47:24+02:00

Hardkernel's ODROID GO handheld game kit is well known for his ability to emulate Gameboy and many more game consoles as also for ports of games like 3D Wolfenstein or Doom.

But with his versatile peripheral hardware like e.g. WiFi, Bluetooth and his expansion port with GPIO pins it allows a wide range of applications. Example applications with external hardware are a Weather Station and an Ultrasonic Distance Meter.

Programming is easy with the Arduino IDE even for novices. A lot of libraries are already available because ORDROID GO is based on the widely used ESP32 platform.

As first application I started developing of a "WiFi/network tool". Currently it can do some network tests and light configuration changes in our network. This is an example screenshot:

screenshot wific application

The Wifi connection works very well even through some walls. This Proof Of Concept is already very useful and I will continue to enhance it.

Workaround for Unity installation issue "not enough space" Frank W. Bergmann /blog/archives/2019/09/20/workaround_for_unity_installation_issue_not_enough_space/index.html 2019-09-20T23:22:47+02:00 2019-09-20T23:22:47+02:00

Some days ago my son told me that he had an issue on installing Unity via UnityHubSetup on Linux: Even with specifying TMPDIR and selecting home partition as installation target the installer did still show the (low) free disk space of the root partition.

There are several pages about this topic but apparently there's no easy solution for this issue. I told him that (assuming that you actually have enough free disk space) the disk space can be "faked" by pre-loading a shared library where the statfs() calls are wrapping the system calls and are "faking" the results.

As Proof-of-Concept and for educational purposes I wrote this fake library and a small shell script to control the Unity installer and the lib. A small issue is that the executable UnityHubSetup.AppImage is actually unpacking some kind of image and then starting the actual installer from this image. As a solution my script will launch UnityHubSetup.AppImage to get the image unpacked, then terminate UnityHubSetup.AppImage and finally start the real installer with the fake library preloaded. That causes the installer window flashing one time and then opening up finally.

Script and fake lib as archive: unityhubsetup-fake-diskspace.tar.gz

Content of the archive: 

$ ls -1 unityhubsetup-fake-diskspace
README
launch-uhs-with-faked-increased-diskspace.sh
wrap_statfs.c
wrap_statfs.so

Some technical details: The actual free space is multiplied with a constant factor... 

$ grep -B3 'resize_factor =' wrap_statfs.c 
/**
 * multiply free space with resize_factor
 */
static const int resize_factor = 16;

... and this is an implementation of a faked call: 

$ fgrep -A9 'int statfs(' wrap_statfs.c 
int statfs(const char *path, struct statfs *buf) {
  int result;
  write(2, MESSAGE_STATFS, (sizeof MESSAGE_STATFS) - 1);
  actual_statfs = dlsym(RTLD_NEXT, "statfs");
  result = actual_statfs(path, buf);
  if (result == 0) {
    SET(buf->f_blocks, buf->f_bfree, buf->f_bavail);
  }
  return result;
}
RHEL 5 dovecot fix for CVE-2019-11500 Frank W. Bergmann /blog/archives/2019/08/31/rhel_5_dovecot_fix_for_cve-2019-11500/index.html 2019-08-31T16:34:26+02:00 2019-08-31T16:34:26+02:00