jowi

I'm running some python scripts on array startup; each script sends an email when it is started, and also when an exception occurs, but also when the program exits, using the atexit exit handler (https://medium.com/better-programming/create-exit-handlers-for-your-python-appl-bc279e796b6b) If i run the python script manually, ctrl-c will indeed make the program send an email that it has been stopped, but if i call the python script from a user-script, and stop the program from user-script, no exit email is being sent, the exit handler doesn't seem to be called. I guess this has to do with the way user-script runs these in the background, is there another way to hook up to some exit function in python so this works?

Just discovered the user scripts plugin. Why no one pointed me to this i don't know, but that is a way nicer way to do these sort of things

I've upgraded to 6.8.3. Also upgraded my 256MB cache SSD to a 1TB WD Black Nvme on a supermicro pci card. I also always had a problem with the supermicro x9scm memory, it has 4 slots, and i could never use the 4th slot, when i add a memory module to it (and yes i am using the correct memory) unraid would not boot or crash after 10-20 minutes. I could never figure out why, but now i found that setting the memory refresh rate to X2 in the BIOS, it actually works, so now i get the full 32GB as well I've also ordered 2x 16TB Seagate EXOS to replace my parity disk and some older 4TB data disks. Will format the new ones as XFS and after that will re-format all other disks tot xfs as well one by one. So, a lot of upgrades. Apart from the mystery shares everything just runs fine. Will keep a close look next couple of days.

I had to migrate from an old version which ran only plugins to a newer one that had dockers and supported vm's. That was quite a step for me, took me a lot of blood sweat & tears to get my 'world' running again on 6.5.3 with dockers etc. After that i did not want to upgrade anymore for a while. When it runs, it runs. For me it's a tool, not a high maintenance project i want to tinker with 24/7.

At some point you could upgrade like every month, and you had to convert this, change that, leave this, stop using that but add this, etc etc... way too much changes and complications, i decided i upgrade to this version and stay with it as long as it does what i need, and just works. And it is still version 6.5 so its not like ancient? Haven’t got round leaving reiserfs like i said, you need lots of empty space to convert and i hadn’t. And again, everything just works/worked fine. I will upgrade to 6.9.2 when it is released, i tested it a few days on a spare laptop, got familiair with it, should be a smooth transition. I just tested disk1 because that was the disk where the spooky mystery shares were created.

Done. unraid-diagnostics-20200720-2025.zip

None of the shares use the cache disk. The 'cache' in my system does NOT act as a cache, just as an SSD. Also i have no mover process etc. Did a reiserfsck on disk1, no errors, also did disk2, just as a test, no errors. Did not test all others (it takes about an hour per disk) but since i found no errors i'm confident the others are ok as well. Restarted the array and the 'pinkpop' mystery share is now gone...

ReiserFS was all we got back in the day... never had an opportunity due to the array being 90% full, to empty a disk, reformat etc... i will replace my 4TB parity and 1 of the oldest disks (both over 70.000 hours on each one) with 2x 16TB disks in the near future, and those will be formatted with xfs. That way i have some empty space and i can do all other disks one by one as well. I've set my ssd as cache drive but i'm not using it as an actual cache, no need for it, i just need a fast disk for vm's, appdata, plugins, nextcloud etc. with newer unraid versions you can do this with non assigned disks? All i know this was not possible in older versions. So that is why i don't have user0. In unraid all shares are set up for which disk(s) the use or span. Why this is not in diagnostics, i dont know...

I've chmod-ded the 4 files to 777, they don't show up in the Finder; if i click on the network share (the left arrow that is) i get an error (below). The files do show up if i go to the actual location /mnt/disk1/music/download/pinkpop (or the user share /mnt/user/music/download/pinkpop). But the share that isn't a share, doesn't contain anything and returns a fault. Also in windows.

Voor mij hoeven NL vertalingen niet. Het wordt altijd krom en onleesbaar. Inderdaad ‘fout’ ipv error, dat gaat dan nog wel, maar ook aanmelden ipv inloggen of login (waarbij ik “aanmelden” zie als “registreren”...) maar echt irritant zijn dingen als “lege aanwijzer uitzondering” (null pointer exception) of ‘er is een uitzondering opgetreden’... (deze zie je letterlijk in nederlandse windows vertalingen...) of ‘voorwerp verwijzing is niet geplaatst op een voorbeeld van een voorwerp’ (jullien weten deze wel...) en meer van dat soort kromme idioterie. Nee, doe mij maar gewoon de engelse versie. We zijn Frankrijk of china niet, wij spreken onze talen wel.

It happened again. I created a new folder inside my /mnt/disk1/music/downloads folder (where 'Music' is an actual unraid share on disk1), named 'pinkpop', copied some files to it, and the folder shows up in osx and windows as an empty (!) user share... i checked all disks, it is not by mistake created on any disk's root so unraid will see it as a new share, it only exists inside the /mnt/disk1/music/downloads folder... I know i'm not running the most recent version of unraid (mine is 6.5.3, when 6.9.2 is out i will update) but surely this should not happen? Is this because i have AFP enabled on unraid? I've read that this is obsolete in newer versions? I have it enabled to access my timemachine folder on an unraid disk...

Weird things happening here... i tried to stop the array to see if that solved the mystery share. But it got stuck stopping the array. It took forever, it got stuck in unmounting user shares, could not stop the array. Rebooted the server. This did not go as usual... Powerdown from unraid got stuk as well, after 90 seconds it started forcing a powerdown. Stuck again. In the end, i just IPMI'd a powerdown... After that, rebooting went strange as well... booting went ok but starting the array took about 10 minutes... Usually starting the array takes seconds. This time it mounted disk 1, that took about a minute, than disk 2... again a minute... up to disk 9. Now it is started... the mystery share is gone. The log is showing message like : UNRAID kernel: REISERFS (device md8): replayed 46 transactions in 39 seconds etc etc Running 6.5.3. Last reboot was 107 days ago. unraid-diagnostics-20200716-1207.zip

When i look at the unraid shares (user shares and disks) from my mac, or a parallels windows vm, or from an unraid windows vm, there is a network folder showing up called 'Rock It' that i did NOT create (again, i did NOT create it) and i can not see it anywhere on my unraid server itself? It is not a defined share... Where does this come from? I can't connect to it, resource can not be found etc. Is this some sort of symlink? How do i get rid of it? I tried listing symlinks in the /mnt/user folder but that also did not show anything?

Ok... so autohotkey is NOT a solution to the actual problem, which is how do you copy data/text from the host TO the (Windows)/VM and paste it IN the (Windows) VM.

Tested beta 25 on a spare laptop today, very nice, One thing i noticed, i chose encrypted btfrs as file system in disk settings, formatted my disk, but it shows up as xfs anyway? Is this an issue or did i do something wrong? Also I did had to enter a passphrase to start the array, i guess this is related to the encrypted disk? But... if i stop and restart the array, it asks for a passphrase again and i can enter anything i want, i doesn’t have to be the same as i entered before? What’s the use of that?

Tested beta 25 on a spare laptop today, very nice, One thing i noticed, i chose encrypted btfrs as file system in disk settings, formatted my disk, but it shows up as xfs anyway? Is this an issue or did i do something wrong? Also I did had to enter a passphrase to start the array, i guess this is related to the encrypted disk? But... if i stop and restart the array, it asks for a passphrase again and i can enter anything i want, i doesn’t have to be the same as i entered before? What’s the use of that?

Yes. I could not get it to work... shft ctrl v just pasts the content of the actual local clipboard, not the vnc menu clipboard.

Doesn't work...

ANY news on this yet?

There has been no comments on any question whatsoever for over a month...

Any comments on this?

Yeah, i know, and i do. But then you have to rewire everything... and there is not much room in the closet. Maybe i just wire it up anyway, so when it is needed, i only have to turn it on, and switch the cable modem connection to the emergency router in stead of the unraid server... but then what happens if i reboot the unraid server and the pfsense vm starts up? Then there are 2 routers with ip address 192.168.1.1... or can i just use e.g. 192.168.1.1 for pfsense and 192.168.1.2 for my failover router, and just disable dhcp on the failover router while pfsense is running? I would like to 'switch' if unraid/pfsense is down for some reason, without a lot of recabling. Just unplug the cablemodem from pfsense, and plug it in the failover router, enable dhcp and get on with it? Is that possible?

I've been running pfsense in a vm on a headless UNRAID server for some time now, and it works great... until the server has an issue. And then you can't do anything... you can't reach pfsense, you can't ssh or even IPMI into unraid... you don't have network, you don't have internet. The only thing you can do is turn the server off HARD and pray that your disks are ok... And IF you are rebooting, you can't IPMI to enter bios or even see bios etc because you can reach the server only after it has booted and pfsense is started... it's a great way get into pfsense, but i'm gonna go the dedicated pfsense hardware route as soon as i can.

Yeah, could be some hardware fault in the specific slot. I've cleaned it with air from an air-can and a anti-static brush, but did not help. Maybe there is a pin just a bit out of place, who knows. Btw, this is an Intel platform as well, the motherboard brand is Supermicro (server board).

I've been using the X9SCM for some years now. In the beginning i added (the absolutely correct) 4x 4GB non-buffered ECC memory, and i noticed that sometimes the board didn't boot (memory errors) or it did boot but unraid crashed after an hour, or a day. After a lot of testing i found out that if i just used the first 3 of the 4 banks (in the order DIMM2A, DIMM2B (the blue ones) DIMM1A) all worked fine, but if i also added memory to the last bank (DIMM1B) the errors reoccured... Fed up with it i decided 12GB was enough at the moment. But now i'm in need of more memory, and recently i bought 4x 8GB modules (again, the correct ones according to the manual) also hoping that this time i could use all four modules, assuming there was some sort of incompatibility or other issue with the 4GB ones. But again, once i add the 4th 8GB module to the last (DIMM1B) slot, the system crashes after some time. Can take an hour, or a day, but it will become unresponsive and the only way is to pull the plug... with 3x 8GB the system runs fine. No problems at all. It is up 24/7 for weeks now. I know this is not a very new or modern board, but i was hoping if this was familiar to someone? Or maybe anyone has an idea what is going on? Is this a known issue with this board? Am i missing some BIOS setting? I mean 24GB in stead of 32GB is more than enough for my purpose at the moment, but i purchased 32GB and i want to see 32GB in unraid's system info!