HyperV replication fails

I have a servers with replication turned on. Had to install windows- and HW-updates. Live migration functioned exceptionally well and updates were a breeze.

But, turning on the replication was not easy. I got an error: “Replication failed”. Went to event log and saw this:

“The description for Event ID 32000 from source Microsoft-Windows-Hyper-V-VMMS cannot be found. Either the component that raises this event is not installed on your local computer or the installation is corrupted. You can install or repair the component on the local computer.

If the event originated on another computer, the display information had to be saved with the event.”

After furious googling, found out that the directory containing the replicas (at replica server) could be missing. Indeed that was the cause!

 

But one have to wonder, why not display any useful error messages(s)? Apple-genaration might not expect anything to fail but with MS, it is more common to have problems than have a success (sarcasm).

 

Windows 10 installation problems

I had major problem on clean install of an Windows 10 Enterprise. I used the mediacreation (from MS) -tool to create an USB stick from Windows Enterprise ISO-image.

The errors include; “We couldn’t create a new partition or locate an existing one” “0x80004005” “windows cannot be installed to this disk”.

Searched for a solution, the BIOS settings did not do anything nor the other tips I encountered. Then, one youtube video contained a small piece of knowledge; the mediacreation tool did not do the right thing, the user had to “burn” the USB stick with a different program.

In this case Rufus: https://rufus.akeo.ie/

When you open the Rufust and put the USB stick in, you can choose GPT type partition, choose and select the OS ISO and Start. After “burning” and booting the computer, the Windows installation was able to find the HDD and proceed with the installation.

 

 

 

 

 

 

 

I had no success on changing the BIOS boot order or choosing the legacy option. I hope this saves you some time digging around!

 

Software deveplopment woes

As a person with a long career at IT, it still manages to surprise how things “progress” in development.

A typical software product and development goes in a cycle; a need – solution building – version 1 product. Subsequent iteration improves the product to version 2 and so on.

Especially in a commerce world, where money is everything, this development cycle is a bit different after a few major releases (and typically many years in development). This means multiple developers have evolved the product and at some point the mess will become so big / technology goes obsolete, so the company decides to re-create from scratch (at least when possible to do so) or there is the totally wrong way; re-creating the product just for the money perspective, changes are made to justify more billing NOT based on user needs.

In the end, this creates a totally new product but perhaps made from ground zero. This can be definitely done in a good way; there is already the existing product with known features, userbase and needs.

Then there is the bad way.

I will use the pgAdmin for example. The release 3 was the “old codebase” and then came release 4 and finally release 4 version 2. One would expect it to contain all the features the previous had, or at least the good things in it andd perhaps some new.

The release 4 was made using a new framework with some new features added. I believe there is a totally new group creating the release 4, it has too many problems compared to the previous one.

And striking problem is the quality assurance.

If you start the program, you already can notice something is terribly wrong. The load time is way too long even with a modern machine. The slowness continues even when the program has started. Then there are all the functional problems; fading windows which disappear or animating windows that don’t need to be animated, you can’t change the layout enough, the layout settings are not remembered, viewing the table contents is a major joke; you can view 100 lines but you can’t change the query, instead you have to open a new query window and type the SQL there (or copy& paste it again). The old version enabled the user to change columns one by one – version 4 supports it too BUT YOU HAVE TO SAVE every change by yourself! What happens when you have more rows that can’t fit on single screen? One would have a scrollbar and user just scrolls down. Not with release 4. You can have some rows displayed (even when selected to show all), then scrolling down stops, loads more and scrollbar moves a little. Not seeing enough? Scroll down, yet another stop and load. Want to go to last page? Yes, you can, but you have to click-wait-scroll-click-wait-scroll-click-wait-scroll-click-wait-scroll. Did I mention the release 4 is slow? It is, fading and animated tree-structures help a lot *sigh*

The most stupidest thing were the fading message(s), you could restore the DB but it had some problem, the error was shown but if you did not watch the restore all the time, the message faded away and the user is left wondering what happened. Want to see how long a long query takes? You can, but you must wait and observe the program, it will display how long the query took and will fade it away if you are not fast enough to notice.

pgAdmin is a free product and does not deserve to be blamed. I am sorry for this, but I can not imagine why these kinds of things are not addressed before the release.

 

Telian “asiakaspalvelu”

Joudun työni puolesta olemaan mm. Telian kanssa tekemisissä ja tänään taas varaventtiili soi kun heidän asiakaspalvelunsa on niin olematon.

Chat-palvelut ovat päivän sana monessa firmassa “säästöjen” takia (oikea säästö tulee vasta jos asiakas kykenee ulkoistamaan itsensä muualle paremman asiakaspalvelun pariin).

Meillä on eräs ongelma ja johon tarvitsin pikaista vastausta. Valitsin chatin, avautui ruutu asiakaspalvelu vaiko vikapalvelu. Valitsin vikapalvelun ja kirjoittelin stoorin valmiiksi nopeuttaakseni asiointia. Lähetys ja tuli ilmoitus että aspa onkin varattu ja kommentteja ei voinutkaan enää laittaa lisää, eikä jonossa voi olla.

Onneksi aavistin tilanteen ennakolta ja kopioin tarinan talteen -> valitsin seuraavaksi asiakaspalvelun. Sieltä tulikin heti että kaikki varattuja ja pitää syöttää asiakastiedot, nimi, puhnro (maatunnuksella, muuten ei kelvannut, eikä osannut sanoa tästä vaatimuksesta), firman Y-tunnus (!!) ja sähköpostiosoite.

KAIKKI nämä tiedot oli jo Telialla tiedossa, koska chat palveluun piti erikseen kirjautua firman tunnuksella! Lisäksi jaksaa mietityttää rohkeneeko Telia olla erilainen ja arvostaa asiakkaitaan riippuen kuinka lypsettävästä firmasta on kyse.

Lähetyksen jälkeen mitään chattia ei tullutkaan, vaan tekstit olivat menneet sähköpostiin ja vastausta ei heti saanutkaan.

Mietin että koitan taas vikapalvelua mutta ainoa vaihtoehto olikin jatkaa 1. chattia (uutta ei saanut valita) ja joka sanoi että kaikki varattuja. Kun kirjautui pois ja takaisin, saattoi chatin taas avata “uutena” mutta kaikki henkilöt edelleen varattuja. Luonnollisesti vika jäi taas selvittämättä.

 

Finnair ja nettipalvelun taso

Minua jaksaa edelleen hämmästyttää isojen firmojen nettisivut mitä suurempi firma, sen surkeammat sivut.

Esimerkkinä viimeisin, lentovaraus Finnairilla, sivut ovat suht hitaat ja jos klikkailet liian nopeasti valintoja, ei se niitä kerkeä tajuta. Lopulta yritän päästä maksamaan ja sivu pukkaa sinnikkäästi “Error 500–Internal Server Error”.

Nyt hetken odottelun ja sivupäivityksien jälkeen maksusivu latautui.

 

Ubuntu 16, Bind and disabling ipv6

I furiously tried to convert bind to operate only on ipv4 addresses but no success.

Ubuntu way is to add-4 to OPTIONS at /etc/default/bind9

-> No success 

After googling I found out the systemd has hijacked the startup and does not use the bind9-config file at all.

A proper script to modify is /lib/systemd/system/bind9.service
..
ExecStart=/usr/sbin/named -f -u bind
->
ExecStart=/usr/sbin/named -4 -f -u bind

Save and

systemctl daemon-reload

service bind9 restart

-> Success

 

noip.com

I had to reinstall one of my servers and had trouble with re-installation of the noip2-package. I was unable to find the configuration file and google search on example no-ip2.conf-file came up empty.

The reason for not finding the example file was simple: the file is binary and contents were like ‘0.0.0.0^@^@^@^@^@^@^@^@^@<^@^^’.

So if you have to reinstall and create the config-file with parameter -C, please note it will take at least 10 minutes for any password change(s) to propage to noip’s servers.

 

Linux 4.4 and Phenom powersave

I tried to make my old Phenom II X4 B50 be more power friendly with the newest ubuntu 16.04 but the CPU’s were stuck at 3.2 GHz.

Cpufreq-set did not report errors, not it did not do the change (never happened me before). I googled a bit and Linux 4.4 is using acpi-cpufreq driver and not the old k6-driver. Also, the power modules are integrated into kernel now, so no unloading the acpi-cpufreq anymore. Cool’n’Quiet was turned on at bios-level so frequency scaling should work.

Luckily I looked at system settings; scaling_min_freq setting and for some reason, it was the max frequency available. By writing the lowest frequency available to the processor, the processors finally went into lower frequencies.

—–

echo 800000 > /sys/devices/system/cpu/cpu0/cpufreq/scaling_min_freq
echo ondemand > /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor
echo 800000 > /sys/devices/system/cpu/cpu1/cpufreq/scaling_min_freq
echo ondemand > /sys/devices/system/cpu/cpu1/cpufreq/scaling_governor
echo 800000 > /sys/devices/system/cpu/cpu2/cpufreq/scaling_min_freq
echo ondemand > /sys/devices/system/cpu/cpu2/cpufreq/scaling_governor
echo 800000 > /sys/devices/system/cpu/cpu3/cpufreq/scaling_min_freq
echo ondemand > /sys/devices/system/cpu/cpu3/cpufreq/scaling_governor

Maissietanolin käyttö bensiinin korvikkeena on vain lisännyt kasvihuonekaasujen päästöjä

”Biopolttoaineet ovat huonompia kuin bensiini”, Michiganin yliopiston tutkija John DeCicco sanoo ilmastoa lämmittävistä päästöistä. ”Ilmastosyiden takia biopolttoaineita suosiva politiikka on nyt todistettu tieteellisesti virheelliseksi.”

Biopolttoaineet lisäävät hiilipäästöjä – ”Kova data vahvistaa pahimmat pelot”

 

Suuryritysten nettisivuista

Olen aiemminkin tästä avautunut että miksi suuryrityksillä on useimmiten ne surkeimmat nettisivut.

Tiedon löytäminen niistä on yleensä jo oma vaikeutensa saati kun joudut ottamaan yhteyttä.

Tänään esimerkkinä IBM, pitäisi tehdä huoltokutsu ja oheinen sivu nostaa verenpainetta:

IBMSelkeästi näemme että järjestelmä on puhelinnumeroa vailla.

Oman profiilin päivityksen jokaisessa kentässä on arvot ja numero ei vain siirry.

*HUOH*

 

 

 

 

** Päivitys **

Kun kirjautui ulos ja takaisin, puhelinnumero päivittyi. Kuitenkin ansaitsee huokauksen kankeudelleen.