Posts by RetroLarry

And that's the weird thing to me... Assuming that KERNEL and SYSTEM is getting bootstrapped correctly, why is the same disk= instruction working when SYSTEM is coming from local (ie SSD), but not working when SYSTEM is coming from NFS?

However, I setup two "important" diskless LE boxes (they're all NUC like and having a SD slot) to boot from a syslinux'd SD, containing KERNEL and SYSTEM and still mounting their HOME from NFS. Also fast and reliable working like iPXE only, so I can live with that.

I'll leave that issue opened for more experiments and comments, if any.

Quote from mglae

Add the required NFSV4 parameter, e.g. ,vers=4.2. The wiki was updated recently.

I'm still surprised that it was working before without, as reported in this thread too.

Thanks, but doesn't work. Tried with 3, 4 and 4.2 for "vers", added to boot= and/or disk=
Server setup is providing a working NFSv3 and v´4.2 for daily use, so this should be out of question.

What I'm still wondering about is the (new) message as shown in pic#1:
mkdir: can’t create directory ’/storage/00012e00500E”: Read-only file system
Sounds like HOME was found and mounted, but is read-only.

The LE in our living room is booting KERNEL and SYSTEM from SSD but is mounting its HOME by NFS. This is still working after the update.
start by syslinux is
LABEL LibreElec 
   MENU LABEL LibreElec 
   LINUX  /KERNEL 
   INITRD /intel-ucode.img 
   APPEND ip=:::::eth0:dhcp boot=/dev/sda1 disk=NFS=192.168.0.3:/srv/nfs4/pxe/libreelec mitigations=off overlay quiet

Every other room in our house is running a diskless LE setup for years. They're booting as documentated in https://wiki.libreelec.tv/configuration/network-boot
Yesterday I've updated KERNEL and SYSTEM on the (PXE) server and now none of the boxes are correctly booting up anymore.
iPXE start in detail is
kernel http://${next-server}/pxe/linux/libreelecx64 
imgargs libreelecx64 ip=:::::eth0:dhcp boot=NFS=${next-server}:/srv/nfs4/pxe/libreelec disk=NFS=${next-server}:/srv/nfs4/pxe/libreelec overlay quiet

When they/one of them is booting up I can see that it is mounting its HOME - or tyring to do so...boot doesn't continue and no file in HOME is touched.

Jan 25 12:33:31 debianpve rpc.mountd[1457]: authenticated mount request from 192.168.2.11:950 for /srv/nfs4/pxe/libreelec (/srv/nfs4/pxe)
Jan 25 12:33:31 debianpve rpc.mountd[1457]: authenticated mount request from 192.168.2.11:857 for /srv/nfs4/pxe/libreelec (/srv/nfs4/pxe)
Jan 25 12:33:31 debianpve rpc.mountd[1457]: authenticated mount request from 192.168.2.11:816 for /srv/nfs4/pxe/libreelec/00012e00800e (/srv/nfs4/pxe)

Now I'm a little confused and looking for help...

The database problem in detail were duplicate entries in table movie/idFile which prevented a full upgrade of the old database.

For anybody having similar issues while upgrading LE: A simple SQL code is showing duplicates:

SELECT idFile, COUNT(*) c FROM movie GROUP BY idFile HAVING c > 1;

Remove all duplicate entries by hand or another SQL command, rerun LE and have a close look at kodi.log how the migrations of the old database is going.

...it was a database problrm. Three LE12 instances are now updated and working.
DB update went from v121 to v131 but has reproducible issues while migrating. Should be worth to open a ticket at Kodi's forum...

Thanks, I've already seen and set it but no difference.

But I've found out that the MySQL (MariaDB) db migration doesn't complete successfully and LE complained about wrong and/or missing tables (the _view ones). I dumped and killed the old db, let LE12 create the empty/new one and imported the old data into the new db - seems to work fine and NFS and SMB are working again. Looks like the above mentioned issue was a simple database error (incomplete or incorrect migration).

Have to check all other machines now in our house, at least two more, to see if issue is gone and I could close that ticket.

Streaming video from http sources seems to work fine, I can watch TV streams from my SAT2IP receiver/tvheadend. But when playing video files from my NAS LE12 just crashes w/o any meaningful entry in log. This happens with all media types (ie avi/mp4/mkv), codecs and with SMB as NFS shares.

Music files are working, though.

Back to latest LE11 and all is working again.

Said in advance: I don't need help. I'm describing my adventerous experience only, but would be happy to get some opinions.

I have 6 identical small HTPCs since years at home, each with it's own TV or TV+5.1 setup and all of them are running LE (OE in the past). All HTPCs are booting/running LE from network and are having a more or less identical setup.

Yesterday I switched from Leia to Matrix with a reserve HTPC (same hardware as the other 6) and a completely new setup/config. All was working fine for hours, so I deployed the working setup to the server.

In the evening I turned on th HTPC in the living room, which is connected to a HDMI-Receiver/home theater, which in turn is connected to a 55" LG TV (55LA7408). The experience was a completely disaster: The HTPC slowed down after relatively exact 5 minutes, eating up it's remaining RAM in further ~30 seconds, throwed an OOM and restarted LE. This then repeated constantly.

Today I was busy with debugging and after hours I've found that the LG TV was still connected to the LAN, but "unaddressed" by DHCP and firewall 'cause of privacy converns (I left the TV connected in the past for further experiments in the future). After disconnecting the LAN cable from the TV this spooky adventure has ended. And this is the interesting thing; When leaving the LG TV switched on and connected to the LAN, all HTPCs in the LAN are having the issue decribed above.

Which now leaves me puzzling is, what has changed in LE/Kodi? Some automagic detection of "Smart Dirt" (can't write the correct word here because of f*ckin' censorship) ? What was the LG doing in the LAN...broadcating to hell?

At the end my realization and advice is: disconnect all these shitty smart things from your network or replace it's software with open source, where applicable.

Anybody else with an overheating Wetek Hub? Mine is a new one, running with LE 9.0.2MR from MicroSD.

As long as it is working, the Wetek Hub runs fine and smooth. Nothing to complain about. The CPU temperature raises up to ~69°C/156°F and after ~2hrs of work the Hub becomes unresponsible and after some more minutes it hangs completely. No response and no access to it's IP. This happens independently of an idle Hub or playing media.

I opened up it's case and placed a small fan on top of the heat sink which is pushing the temperature down to 46°C/115°F. No unresponsible Hub until now.

Maybe I print a new case for the Hub with openings for the fan, but I'm wondering if anybody else experience an overheating?