Lying half awake in bed this morning, heard a “beep”. Damn. Grid power failure? The alarm clock
still showed the correct time. Out into the garage to look at the inverter. Yes, power
out. For once the inverter didn't interrupt house power.
The outage was fortunately brief. Here a timeline:
05:42:18
Power fails
05:44:32
Power restored
05:47
Power fail message from Powercor, estimate 8:30
05:52
Power restore message from Powercor
I suppose that the messages from Powercor are marginally useful
if it's a longer failure, but clearly they suffer badly from latency, and their restoration
estimates are preselected guesses.
The bad news: lagune failed again! I wonder whether I should change the
hardware.
While attending to the power failure, discovered that almost no mail had come in overnight.
That's a sure sign of either a full file system on mail.lemis.com, or
that named has crashed. This time it was the latter. And a couple of hours later it
crashed again! Put a start command in /etc/crontab, and marvelled that it was
restarted after crash at 9:07, 11:07, 12:07, 13:07 and 14:07 UTC (the 07 is because that's
what I put in crontab, once an hour at 7 minutes past the hour). And then it was
quiet. Why? What is killing it?
A question on the Unix Heritage Society list
today: anybody have the CD that came with “Panic!”, a book about Sun crash dump analysis. I
have the book. Does it have the CD? Yes, and it's unopened:
Mar 27 15:39:33 hydra kernel: (cd0:ahcich3:0:0:0): READ(10). CDB: 28 00 00 04 be 23 00 00 01 00
Mar 27 15:39:33 hydra kernel: (cd0:ahcich3:0:0:0): CAM status: SCSI Status Error
Mar 27 15:39:33 hydra kernel: (cd0:ahcich3:0:0:0): SCSI status: Check Condition
Mar 27 15:39:33 hydra kernel: (cd0:ahcich3:0:0:0): SCSI sense: ILLEGAL REQUEST asc:21,0 (Logical block address
out of range)
Mar 27 15:39:33 hydra kernel: (cd0:ahcich3:0:0:0): Error 22, Unretryable error
Mar 27 15:39:33 hydra kernel: (cd0:ahcich3:0:0:0): cddone: got error 0x16 back
What's wrong there? This is a brand new CD. And I couldn't get the CD out of the drive
until it had finished its retries, over 7 minutes later at 15:46:54. Damn. One of the
reasons I chose this form factor was so that I could put a CD/DVD in it. And I'm not sure
that I have ever used it in the 2½ years since I got the machine. Is the drive defective?
Tried other systems. To my surprise, eureoa (Microsoft “Windows” 7) mounted it and reported no problems, but what can I
do with Microsoft? eureka was non-responsive: the DVD drive was disconnected,
probably because of cable limitations. teevee also didn't want to know,
but tiwi did, and it showed the same problem as on hydra. And the reason for
the message in the first place was because the submitter had also had a defective CD. Maybe
they all were.
But why could Microsoft mount it? Tried it in tiwi. To my surprise I was able to
mount it and read the data on it. I wonder if it's correct: I would need an old Sun machine
to check properly, and I gave my last one away some time ago. But that could explain the
Microsoft acceptance: it presumably just didn't check.
For reasons I still don't understand, tried again to access remote file systems from
Android today. It's worse than pulling teeth. This is castrated Linux, after all: why don't they
offer proper network software? After eliminating the paid apps, found Kodi, recommended by
Google Gemini.
What an amazing app! It's not documented, of course, and some of the sample images are in
Dutch. I can't work out how to navigate it. But the Gemini instuctions told me to navigate
to places that I couldn't find. And when I finally found what appeared to be the correct
place, I had to type in a pathname (no way past that, I suppose), but on an even
tinier and more horrible “keyboard” than standard. Finally I got the name typed in, pressed
a wrong button, and it was all gone again.
Finally I had it a little faster (hint: aim for the “keyboard” row above what you would
expect), but it didn't want to know.
FOOL! You need to use the standard path name specifications.
Not hydra:/src. It should be nfs://192.109.197.129/src! As the twins
say:
Kodi uses the standard NFS URI format to address shares. If the "Browse" function isn't
automatically finding your server, you must enter the path manually using this syntax:
The Standard Format
The URI should look like this:
nfs://[IP_ADDRESS]/[EXPORT_PATH]/
Example:
If your FreeBSD server is at 192.168.1.50 and your export in /etc/exports is
/usr/home/media, the name in Kodi should be:
nfs://192.168.1.50/usr/home/media/
Amazing! For over 30 years I have been using non-standard path names! I wonder if it will
accept names. But first just get the bloody thing to work. That, too, didn't work:
I think that these messages are related. Kodi politely overwrites the most important part
of the screen with an incorrect guess at the cause. This time, though, I got a more usable
error message, from a more usable system:
Mar 27 16:51:32 hydra mountd[1931]: mount request from 192.109.197.242 from unprivileged port
Mar 27 16:52:07 hydra syslogd: last message repeated 4 times
OK, I can live with that. At least I had an error message that I could use. And in the
meantime the twins mentioned a “Browse” function. Mañana.
But why doesn't Android have a proper network stack?
Now where did that come from? And how do I find it in a more normal format?
There was also yet another message, about Nano Banana (a name that I thought they were
trying to remove), which I must read if I can ever find it again.
Schweinshaxe again for
dinner this evening. Last time things were good, but the skin was a little unevenly browned. OK, how
about in an air fryer? The “bathtub” seemed to
be a good idea, but then it occurred to me that maybe doing it in the spit would be better,
for which I needed the “hair dryer”.
But how do you position the skewer? One end is almost only bone:
Contrary to my expectations, there was no difficulty in inserting the skewer, but my first
attempt came out on the same side of the bone, making it very lopsided. It has to go on the
other side (second photo);
The Hax'n weighed 1.3 kg, which according to the instructions should have taken 65 minutes
at 240°. For some reason I chose 220° instead, but even so, it looked pretty well cooked
after 25 minutes:
We've been using LED light bulbs for over 10 years, and the ones between kitchen and lounge
room have been there for 5. But lately one of them has changed behaviour:
Why is that? In case of doubt, it means that the lamp is dying, but why should it then stay
on for longer without power? Is there something in the globe that turns off the light when
the power goes away?
This page contains (roughly) yesterday's and today's entries. I have
a horror of reverse chronological documents, so
all my diary entries are chronological. This page normally contains the last two days,
but if I fall behind it may contain more. You can find older entries in
the archive. Note that I often update a diary entry
a day or two after I write it.
Do you have a comment about something I have written? This is a diary, not a
“blog”, and there is
deliberately no provision for directly adding comments. It's also not a vehicle
for third-party content. But I welcome feedback and try to
reply to all messages I receive. See the diary overview for more details. If you do
send me a message relating to something I have written, please indicate
whether you'd prefer me not to mention your name. Otherwise
I'll assume that it's OK to do so.
