thanks, i have the flashmemory plugin installed. but i haven't had any unnormal shutdowns.
Anyway because of the remark to search the forum again, i've found a command which reinstalled nginx and stuff.
Sorry, didn't find the command and or link to the forum posting. However, since then the server boots again without problems
really? i did, thanks doing again
dunno how i come to this and how to debug this, but this is happening
- server reboot
- omv starts everything but webinterface
- nginx -t shows that the nginx folder in var/log is missing
- mkdir /var/log/ngnix solves the nginx -t error
- systemctl start nginx - starts fine
- webinterface is there - login works
as soon as i reboot, the thing starts at the beginning of my list and webinterface doesn't launch.
dunno what i did......
of course i did. using kvm plugin every day
not expecting anything special. Just recognized my central logging infra notified me about this log not written to sine last May. If this is by design, so, thanks for clarification.
okay, then the backup logs are there, but in /var/log i have a entry of "6. Mai 2024 openmediavault-kvm.log" which is the only one.....
just recorgnized that within my working KVM environment the logging had stooped back in may of this year. 
also KVM Backup logging has stopped. KVM Backup itself is working
How can i reactivate this?
Other logging like messages or syslog is working....
back to smartctl and drivedb.h as local database.
As my understand the database should provide HDD information and modified commands to access those HDDs, like the -d megaraid.
Exactly with those details commands i'm struggling, where to put those
i guess then this gets executed accordingly:
/openmediavault/engined/rpc/smart.inc: $cmd = new \OMV\System\Process("smartctl", $cmdArgs);
I've modified a downloaded drivedb.h and stored it in /etc/drivedb.h and restarted smartmontools service. The entry change was my drive:
{ "HGST Ultrastar He10", // tested with HGST HUH7210100ALE600/0F27452
"HGST HUH7210(08|10)AL[EN]60[014]",
"", "", "-d megaraid,04"
//"-v 22,raw48,Helium_Level"
},
{ "Western Digital Ultrastar (He10/12)", // WD white label, tested with
// WDC WD80EMAZ-00WJTA0/83.H0A83 (Easystore 0x1058:0x25fb),
// WDC WD80EZAZ-11TDBA0/83.H0A83, WDC WD100EMAZ-00WJTA0/83.H0A83,
// WDC WD100EZAZ-11TDBA0/83.H0A83, WDC WD120EMAZ-11BLFA0/81.00A81
// WDC WD140EDFZ-11A0VA0/81.00A81 (Easystore 0x1058:0x25fb)
// WDC WD140EDGZ-11B2DA2/85.00A85, WDC WD140EDGZ-11B1PA0/85.00A85
// WDC WD120EDAZ-11F3RA0/81.00A81, WDC WD80EDAZ-11TA3A0/81.00A81
// WDC WD40EDAZ-11SLVB0/80.00A80
"WDC WD(40EDA|(80|100|120|140)E([MZ]A|DA|DF|DG))Z-.*",
"", "", ""
//"-v 22,raw48,Helium_Level" // not: WD80EDAZ, WD40EDAZ
},
as you can see, i've added for a test a "-d megaraid,04"
but this doesn't lead to changes in the SMART area of omv web ui.
It still comes up with an "use smartctl with -d" to show stuff here....
when i create a file
/etc/smart_drivedb.h with only the above entries then nothing changes but syslog shows
smartd[159304]: /etc/smart_drivedb.h(3): Syntax error in preset option string
do you see my error?
see attachment. Aren't those messages "good news", aka accepted by smartmontools?
if so, then i have an easier (for me) solution found by modifying /etc/smartd.conf
added lines in this way:
/dev/sda -a -d megaraid,00 -m email@address -M exec /usr/share/smartmontools/smartd-runner
...oh, omv overwrites the file... thats bad...
yes, i've tried to unterstand the custom local database, need to look deeper into the documentation to get used to the needed format/information for my drives.
so, the migration is finished. Out of the many minor issues, two major issues stand out at the end.
1.) deleting the SW Raid in OMV webui is difficult, cause every reference to it must go. So the last task before deletion is unmounting the filesystem. Then a delete is possible. Really? No, but.... there is no reference anymore... why the heck.... damnit! Make a reboot after unmount, then it works.
2.( as with a new HW RAID all references in the /srv/ path (the dev-1234.... references changes. Thats the main problem. Every hardlink in scripts, docker compose files, VM xmls must be changed.
So the learning here is: work with the symlinks plugin as much as possible and created references like /mnt/docker for example
Makes your life easier at the end
back in 2022 votdev referenced this solution
for people with HW RAID controllers,.
I've tried the same with OMV7. But this is not working
the valid command smartctl -a /dev/sda -d megaraid,00
..and referencing <devicetype>megaraid,00</devicetype>
in config.xml and restarting openmediavault-engined service
doesn't lead to the expected results. Are there other solutions for OMV7?
just for the sake of completion. What are the values of this "default network"?
ah okay, then i had no networks in kvm plugin from the beginning. understood
your screenshot is not clear to me, with the mentioning at the same time that you create a host bridge.
I have the same (see xml snippet) br0 for 1GB network and br1 for 10GB, but you've not referenced these networks in your screenshot
That is a VM NIC using the host's bridge (not a kvm network). Since VMs just reference networks by name, you can just create a new network in the Networks tab of the KVM plugin. I'm still confused how you lost the networks.
i've played around with bonding and afterwards recreated the br0 interface and everything in the omv network area, but
nothing in kvm network area.
which network should i create then in kvm?
the net0, the br0? i'm a little bit confused about mac address and ip range /24 and so one there.....
Yes i've not deleted the networks. In the mentioned folder i have only an autostart folder, also nothing in the backup.
in my xml file of a vm for example i have this:
<interface type='bridge'> <mac address='52:54:00:bc:44:02'/> <source bridge='br0'/> <target dev='vnet2'/> <model type='e1000'/> <alias name='net0'/>
can this be used? in the kmv web gui?
found it in backup....
...but now i've discovered that my networks in the omv plugin are gone, no default, no nothing.... oh well, can they be recreated from a backup config?
