Everything posted by MAM59
-
In case you still are interested... in the mean time I have created a new share "B" on that drive. The share was automatically exclusive. Then I have a hard time shuffling over the Data from Backup to B. The "mv" command of unraid is as stupid as it can be. It is a "copy then delete" thing, safe but wasting time. The real "mv" sees "ooh! source an target are on the same filesystem! i can just rename the files instead of copying terrabytes of data!" This is a serious time killer and without any reason. Why not use the usual "mv" instead??? Anyway, still a few hundred gigs to go...
-
Tried both already. no go To make it clear again (maybe my english is not good enough): The "Backup" share existed for a long time on the (UNRAID) Array. Now I have put in a new disk and made it the pool "backpool". Then I have moved over all the files from the various "Backup" folders on the array disk (erasing the empty folders afterwards). I thought that "exclusive share" would be switched on automatically now like it did for many shares before. But no, it remained un-exclusive. I rebootet, still un-exclusive (to make sure that there are no pending connections to the share) Then I wrote the question here 🙂
-
hit CTRL-F5! the Forum is not really stable currently. the file is there (1st it failed but later on it uploaded correctly (the forum says... :() )) )
-
tnx for your help. Unless I am blind (almost 🙂 ) there is no 2nd folder left But see yourself... The share I am looking for is called "Backup" (yeah, not very innovative I know 🙂 ) f-diagnostics-20250407-1414.zip
-
sure, many other shares are exclusive already. just this one does not love me 🙂
-
Long story short: I moved all the split-among-drives-in-the-array data of a certain share to a new, standalone pool disk. Then I've edited the share config to just use this pool (and 2ndary=none). There are no remaining empty folders elswhere on any other drive. But still it does not turn on "exclusive share" and creates a symlink. Instead the folder is still managed by FUSE... How to get rid of FUSE here ? (To overcome it I plan to create a new share on that disk, move over all the files and then delete the old share. Later on renaming the new one to the old name. But there must be an easier way...)
-
Since a few versions (dunno which one anymore, sorry) I notice an evil problem. When trying to stop the array (or to reboot or something), UNRAID hangs around and sits waiting for Apr 7 09:41:57 F emhttpd: shcmd (220): /usr/sbin/zpool export -f vm Apr 7 09:41:57 F root: cannot unmount '/mnt/vm/system': pool or dataset is busy Apr 7 09:41:57 F emhttpd: shcmd (220): exit status: 1 I noticed that UNRAID is unable to shut down the (one and only) VM "Windows Server 2025"... Ok, so this might be a bit too new for poor QEMU, but if I stop the VM via GUI (before or after the reboot or stop-array), it shuts down and the halted task continue. So I guess there is something wrong with the stop-array script not finding the correct words when talking to the VM manager??? What can be done manually should also work automatically Apr 7 09:42:17 F emhttpd: Unmounting disks... Apr 7 09:42:17 F emhttpd: shcmd (224): /usr/sbin/zpool export -f vm Apr 7 09:42:17 F root: cannot unmount '/mnt/vm/system': pool or dataset is busy Apr 7 09:42:17 F emhttpd: shcmd (224): exit status: 1 Apr 7 09:42:17 F emhttpd: Retry unmounting disk share(s)... Apr 7 09:42:20 F rc.local_shutdown: Stopping emhttpd Apr 7 09:42:20 F rc.local_shutdown: /usr/local/sbin/emhttp stop Apr 7 09:42:20 F emhttp: Stopping web services... Apr 7 09:42:20 F rc.nginx: Stopping Nginx server daemon gracefully...
-
Ach DU warst das 🙂 "...und das ist auch gut so", wie WoWi sagen würde... Die Kacheln waren ein großer Schritt in die falsche Richtung. Niemand will Händifeeling auf dem Desktop, das war eine wirkungsvolle Arbeitsverhinderungsmethode. Die Rolle rückwärts bei 11 hat gut getan- Zumindest für die Leute, die mehr als ein Fenster gleichzeitig offenhaben wollen.. Aber egal, das ist Geschmackssache (aber die Mehrheit teilte wohl nicht Deinen Geschmack 😁)
-
I know "thou shalt not move something complicated", but of course I have tried (to move "docker" for instance) and had to learn it the hard way that mover is not a very bright program (yet?) When it comes to a named pipe for instance, it does not complain, it does not skip, it just hangs around until the cows come home (or if it has been hit with a kill command). Next run, same file, same hang 😞 It should not be very complicated to teach mover to skip such a file (and complain loudly) but continue the run else. So pls Limetech: FIX IT!
-
If I click around in the disks I eventually catch this error: Apr 6 08:54:32 F nginx: 2025/04/06 08:54:32 [error] 11208#11208: *137071 open() "/usr/local/emhttp/plugins/dynamix.docker.manager/styles/style-azure.css" failed (2: No such file or directory) while sending to client, client: 192.168.0.26, server: , request: "GET /plugins/dynamix.docker.manager/styles/style-azure.css?v=autov_fileDoesntExist HTTP/1.1", host: "192.168.0.4:800", referrer: "http://192.168.0.4:800/Main" Seems that style-azure.css is missing, no idea what it is good for (ok, its just a style sheet, so no missing functionality just "not shown as wanted by the author". Still a bit annoying).
-
Warnung: Ist nur ne Vermutung (weil sonst kein Fehler in dieser Richtung irgendwo aufgetreten ist)! Aber um das W11 Update kommste eh nicht mehr lange rum. Ich versteh auch nicht wirklich, was die Leute alles zu meckern haben (ausser die Cloud und KI Gängelei die keiner braucht). Bin kein großer Fan von Microsoft, aber funktionieren tut es schon auf vernünftiger Hardware (manche stricken sich ja was esotherisches zusammen und übertakten solange, bis der Chip in der CPU explodiert)
-
Na ja, mit so veralteten Systemen kann es dann schon mal zu Komplikationen führen... Microsoft hat bei W11 reichlich am SMB LAN gewerkelt und UNRAID (bzw. SAMBA) hat entsprechend nachgezogen. Vielleicht ist UNRAID 7 dann doch keine gute Idee für Dich?
-
I know "thou shalt not move something complicated", but of course I have tried (to move "docker" for instance) and had to learn it the hard way that mover is not a very bright program (yet?) When it comes to a named pipe for instance, it does not complain, it does not skip, it just hangs around until the cows come home (or if it has been hit with a kill command). Next run, same file, same hang 😞 It should not be very complicated to teach mover to skip such a file (and complain loudly) but continue the run else. So pls Limetech: FIX IT!
-
Fangen wird doch mal damit an über welche Windows Version wir hier eigentlich reden???
-
Nimm lieber n Raspi 4 oder 5 mit HA-OS und opfere die 5W. HA auf UNRAID funktioniert zwar, aber durch VM / Docker Ausfälle z.b. durch Backups, sind oft die Statistiken für den laufenden Tag fehlerhaft (z.B. beim Überwachen von Solaranlagen kommen da schon "merkwürdige Werte" vor wenn er alle 5min die Verbräuche buchen will, HA aber gerade gestoppt ist).
-
Jein 🙂 Schuld bist wohl Du selber. Windows hält immer nur EINE Verbindung aufrecht, wenn Du also mit Name angefangen hast, bleib dabei. Oder fang mit IP an und bleib dabei. Niemals mischen!!! Ausserdem gibt es noch reichlich Variationen die auch alle zum Mislingen beitragen können (Host mit Domäne, Username mit Domäne usw.) Es ist also ratsam sich für eine Schreibweise zu entscheiden und immer dabei zu bleiben. Die Schreibrechte entscheidet Unraid anhand des Usernamens. Und der kann ja wie erwähnt auch differieren...
-
Schon ein Problem von Windows... In den Anmeldedaten hast Du Deinen Benutzernamen/Passwort hinterlegt unter \\HOSTNAME\SHARENAME. Der gilt aber nicht, wenn Du per \\IP-ADRESSE-DES-HOSTS\SHARENAME zugreifst. Hierfür musst Du bei Windows Anmeldeinformationen einen zweiten Eintrag mit dem anderen Format hinterlegen (Windows ist blöd. Es merkt nicht, das NAME und IP denselben Rechner meinen. Der Name wird zwar per DNS oder anderen Methoden aufgelöst, die Anmeldedaten werden aber nur mit dem Originalstring verglichen)
-
The GUI is a bit confusing (correct, but not very intuitive). Many options are not available unless you fulfil some (at first) unknown prerequisites. Thats good because it keeps you from picking uttlerly wrong combinations, but also bad for a newbie that does not know what is possible. You have to use "cache" as primary storage for that share. THEN magically the settings for "secondary storage" will appear. Select "array" here. This means, new files first go to "cache", and later are moved to "array" (just what you were looking for). (Note "cache" is just a disk, not a function. Actually it is a "pool drive". For historic reasons the first pool created is automagically named "cache". But you can use ANY drive for primary (and secondary) storage. Only if you use "array" as primary, no option for secondary will be offered again)
-
ooh, sorry, my fault... its /boot/extra (without "s" at the end). But I see from your listing that you do not have this either. You have the non-standard "packages" folder. whats in there??? (dunno where it may belong too) AAAH! just looked it up... its a leftover from the not-anymore-existing plugin "nerd-tools". Obviously it is still honored at reboot time. So I would guess you will find your python here. Clean it up and reboot (also clean up everything else in this folder, because nerd-tools are gone and no updates will happen anymore. If there is "needed" stuff you have to find a new way to install it (like python which now has its own plugin))
-
as already been told... look into the boot/extras folder. If there is something like "python", delete it and REBOOT. This should get you rid of "the old" python.,.. (but be aware that "python for Unraid" will do just the same, download python and populate /boot/extras. You wont get any further with this. Better read the Instructions for this strange and unknown "Docmost". They should tell you WHERE (and for which version) they are looking for python...)
-
orico ns1066 scheint ein externes (USB) Dock für 3.5" Platten zu sein, also so Marke "dumm wie Brot". Ich befürchte, er hat beim "Kopier"vorgang immer eine Platte nach der anderen reingeschoben, in der Hoffnung, dass es da einfach weitergeht. Deshalb wohl die NTFS/extFAT Mischung. Aber, ich vermute stark, DIE DATEN SIND NOCH DA UND HEILE! Er wird wohl die Kopien "normal" in einen PC eingebaut haben, da erscheinen sie dann leer. Diese USB Controller (hier NS1066) sind ja berüchtigt, eine eigene Partitionsstruktur (wie RAID Controller) anzulegen. Da sind die Platten dann auch nur noch in diesem Gehäuse lesbar. Steck mal sonne Platte wieder in dieses Orico Teufelsdingen rein und guck nach, ob sie immer noch "leer" ist...
-
You should configure the printer NOT to use WINS. Its outdated, obsolete and unlikely that there will still be somewhere a server around somewhere. I would not call this an error message, its more a warning to take care about an misconfigured device...
-
Unlikely. What about logging off and then on again?
-
there was another line in your logs that could point to trouble: Mar 21 16:05:26 SAN01 kernel: bond2: Warning: No 802.3ad response from the link partner for any adapters in the bond This means, the connected switch does not honor the bond! Look into it's config, try different bond modes (not all are supported by every vendor) or even drop down to "active backup" (which does not need switch support but also does not aggregate anything too). BTW... are you sure you know what you want? Even with a working 802.3ad bond, there will not be any speedup because every connection always just uses ONE card. Many connections maybe split among cards, but a single connection always only uses ONE of them. So in most cases "bonding" is nothing people really need or what pleases them...
-
Don't worry, I was not sure either (just remembered it faintly). That's why I have tried it out again. I guess I have stumbled upon it whilst playing around with 7.0 beta or so. Maybe it was "added" unintentionally as a side effect in the GUI ? I only mentioned it becaused he has asked for a method that does not set any pools at risk. And this way, only array drives can be affected.
