News archive March 2021 - Koos van den Hout

I recently wanted to do some serious cycling to improve my mood and raise my maximum distance per day again. So I found a day off and set a goal of riding more than 100 kilometers. With a bit of planning on the map I decided that Utrecht - Hilversum - Bussum - Almere - Nijkerk - Utrecht was a good way to get about 105 kilometers cycling.

In the end the odometer stopped at 112.53 kilometers. And I do feel better.

With some alerts set to get the last of the Special Event Station series to celebrate the 200th anniversary of Hellenic War of Independence against Ottoman Turks in the log I now have the full set: at least one contact with each of the special event stations.

Which means the website will generate a nice digital certificate for me which I could print out and hang on the door of the room where I have my radio setup. But that door is already filled so I'll just keep the digital certificate and leave it at that.

It was fun chasing them! My thanks to the organization behind this.

Making a video about my new paddle is one thing, actually using it with the radio is another. I have seen radio amateurs buy expensive morse gear and finding out that learning morse is hard.

I connected the paddle to the radio via the nanokeyer I built and called CQ in a part of the 20 meter band where I expect other users with slow speed.

After one CQ I got an answer from PA5ABW Ab. The same person who taught me morse code!

For a while I had a notification set for someone selling a morse paddle. Finally one came along at a reasonable price so I bought it.

And.. I mentioned this detail to some people at work. Who had an idea of what a morse key is, but didn't know about morse paddles. So with my big mouth I said "I'll make a video about it". This was triggered by the fact that I recently learned about OpenShot non-linear video editor which is available for Linux too.

So I created a video. And found out making a video of 30 seconds is a lot more work than 30 seconds. I watched some tutorial videos about OpenShot first and thought about what I wanted to show. I haven't added spoken comments because I didn't feel like doing those too.

The video isn't great, I can see several beginner mistakes. But I get the point across of what a paddle does. There is a continuity problem because I used sunlight. Which isn't very constant. And I made several clips because I didn't think I would get everything I wanted to show right. But now there are changes in light and a bit in camera angle, even with using a tripod.

And our neighbours were busy hammering indoors, so that can be heard too.

It is always good to have a bit of luck and get a contact with a new country. This evening I saw a call from Hong Kong pop up on my screen with FT8 traffic and made the contact with a bit of a hickup since it was hard for me to receive the transmissions. The signal report showed that my signal made it across easier, so I had confidence and the contact was made.

After that I saw a station from Ghana, which had more trouble decoding my signals, but after a few tries that contact was valid too. Ghana is not a completely new country for me, but it was new on the 40 meter band.

Now to wait for digital confirmation (both show they use Logbook Of The World) and see if I can get a QSL card.

Update: I just noticed I didn't write about a few new entities from recent months. In February I also got Anguilla in the log (an island in the Carribean) and confirmed. This was a case of turning on the radio on a non-standard time and seeing a new country and getting the contact. In March I saw notifications for activity from the UK bases on Cyprus (which are two British overseas territories housing military bases because of the strategic location of Cyprus) which I have been chasing for a while and the contact was made.

Update: All contacts mentioned above confirmed.

After the recent work on updating the TLS settings for the webservers at home there was one element missing: TLSv1.3 support.

This needed an upgrade of openssl and the 'easy' way to get there was a full upgrade of the server running the external facing proxy. So I took that step yesterday evening. Made a snapshot first and started upgrading devuan ascii to beowulf.

After the update a lot of things were broken: I defined a non-standard location for bind9 logging and AppArmor disagreed. Without a working nameserver a lot of stuff breaks internally! So after managing to get on the upgraded system with console I changed the AppArmor rules to allow it. After that things started again.

For the next time I manage to break the resolving nameserver: I should remember that avahi/multicast dns works on most systems even when DNS resolving fails. I checked and I can use .local names to get to the right equipment.

After checking how everything is running for about a day I threw out the old snapshot.

As a number of years before I participated in the EA PSK63 Contest 2021. This is a contest organized by the Spanish Amateur Radio Club Unión de Radioaficionados Españoles and I appreciate their work in this and other contests.

Contacts were made Saturday afternoon/evening and Sunday morning. I decided to go for both 20 and 40 meter band to improve my contest results.

In the end I made 148 contacts, 58 on the 20 meter band and 90 on the 40 meter band. To my surprise when I started Sunday morning there was very little activity on the 20 meter band, but the 40 meter band was already filled with noise, probably from nearby solar power installations. With a bit of timing and luck I could work around the noise peaks and make contacts with the stronger stations. Later in the morning there was a lot more activity on the 20 meter band and new stations rolled in.

It was good to see a lot of to me new Spanish callsigns in this contest. I guess amateur radio in general and contesting has grown in Spain.

I needed a virtual machine with ubuntu so I did the base installation and also configured unattended-upgrades and sendmail to get the results. But I noticed after a while I never saw any mail from that machine.

Problem soon found:

mailer=relay, pri=30131, relay=postbode.idefix.net. [82.95.196.202], dsn=4.0.0, stat=Deferred: Connection timed out with postbode.idefix.net.

The machine wasn't even trying to reach the mailserver over IPv6! On the internal network with servers it will fail over IPv4 because of the portforwarding rule for the port from the outside IP to the mailserver but I never expected an internal machine to try IPv4.

Somehow this seems default for sendmail 8.15.2 in Ubuntu 20.04. I could find someone else asking this: No IPv6 outbound from Sendmail starting with 20.04 but no answers how/why.

At first I suspected systemd-resolved as the old saying goes that all sendmail problems are caused by DNS. But disabling that didn't fix the problem.

I now have the IPv6 address hardcoded in the configuration, that works.

dnl FEATURE(`msp', `postbode.idefix.net', `25')
FEATURE(`msp', `[2001:980:14ca:1::23]', `25')

I also found out the option ResolverOptions=+WorkAroundBrokenAAAA was set but not causing this.

A fun bit of amateur radio is the possibility to get awards for making contacts. There are awards like the American Radio Relay League DX century club for making contacts with at least 100 'countries' or the awards linked to summits on the air which are awards for making contacts from or to a number of mountain tops.

There are also awards for making contacts with special event stations. Currently there are several special event stations on the air celebrating 200 years of Greek independence. So far I have found two award options linked to those stations, but there may be more.

Special event stations are usually linked to a special event, such as this historic event for Greece. It is a way to notice things in history or other events that you normally wouldn't notice.

Special Event Station (SES) series as SX9A, SX8A, SX7A, SX6A, SX5A, SX4A, SX3A, SX2A and SX1A to celebrate the 200th anniversary of Hellenic War of Independence against Ottoman Turks. and The Greek Revolution of 1821 We celebrate the 200th anniversary of Hellenic War of Independence Special Event Call 01-31 March 2021 and 200 Years of Greek Independence Award.

It's a good thing they all have websites where you can see your progress and download the digital awards.

Looking at the newest graphs I created with grafana of system statistics I noticed the available entropy was still getting dangerously low from time to time on the system that runs the home server. For some reason this system has no available hardware random number generator. Even after the earlier changes to add more sources of randomness it was sometimes dropping low, especially during dnssec signing operations.

This does mean that the encryption processes for TLS in the webservers may also get delayed. Which is really not what I want.

Time to update settings on randomsound and haveged: I want a minimum of 2048 bits of available entropy. Sofar, this seems to have the desired effect.

I'm currently following the course The Best TLS and PKI Training Course in the World and learning even more about the workings of encryption, TLS and certificates.

One of the things I learned is to balance security with performance. And I directly used this new insight on my own webservers. The connection which brought you this page from https://idefix.net/ is still encrypted but I saved a few milliseconds on the encrypted setup by switching from a big (4096 bit) RSA private key to a 384 bit ECDSA key which are comparable in cryptographic strength. But the calculations with the ECDSA key are less CPU intense. And yes, I have statistics on page loading times before and after the changeover of the key.

It was a good moment to change private keys anyway, the old keys were more than a year old.

This is one of those areas where I like having my knowledge hands-on. Actually understanding what is happening and why.

For years and years I have been using rrdtool to gather and graph statistics at home. I started gathering home temperatures around 2008 but I see NTP statistics gathering from 2003 and my last mrtg graphs were created in October 2002. So that suggests I've been using rrdtool since that date.

Anyway, I'm looking at newer options. After some asking around I installed influxdb and started gathering data. I adjusted some of my data gathering scripts around rrdtool to also put the data in influxdb.

The easiest data to gather and graph was the load average, available entropy and number of processes for a number of systems at home. So that dashboard has been built and allows selection of the wanted computer.

My first conclusion is just collecting data and thinking what kind of graphs to create later is a lot easier with influxdb. With rrdtool the round robin database is designed around the graphs you want. In this case I just start gathering data and when data has come in start playing with possible graphs from that data.

The next challenge is to set the rules for maintaining the old data. One of the triggers to look at other options was that I was at the end of a nearly 11-year cycle of stored temperatures in rrdtool, and I wanted to keep that history if possible.

I don't have to keep every measurement forever, but with storage being cheap I think I will keep daily averages forever when this is 'production'.