-
nfs to unraid + neovim = "E1509: Error occurred when reading or writing extended attribute"
Getting the error with neovim: ``` E1509: Error occurred when reading or writing extended attribute ``` When I try and save a file. This is causing my vscode to have issues (runs off of a neovim plugin). Same issues with neovim from multiple clients to same nfs share(same 10.0 version; with no plugins loading `--clean`). I do NOT get this issue with neovim 9.2. Other editors (vim, nano) don't return this error. HOWEVER This error does NOT occur on my local drive, nor directly on unraid server (though it's regular vim installed). Other non-unraid nfs mounts seem to work fine. As you can see, I'd almost peg this as a neovim issue exclusively, but everything works fine until I am trying to save on my unraid nfs share. Is there something I can look into here. Is there something in unraid that will allow me to inspect the extended attributes to try and investigate this further? ```fstab #192.168.1.25:/mnt/user/<user. /mnt/bebop_<user> nfs rw,noatime,x-systemd.automount,rsize=131072,wsize=131072 0 0 192.168.1.25:/mnt/user/<user>/ /mnt/bebop_<user> nfs rw,noatime,x-systemd.automount 0 0 ```
-
WebUI too many login attempts
To increase the number of attempts or lockout length edit `/vim/security/faillock.conf` (you'll need to restart or restart services)
-
53kl2s started following UI/NFS no longer supports FQDN mounts
-
UI/NFS no longer supports FQDN mounts
I used to be able to use SPIKE.HOME.SOMENAME.ORG_docker In MAIN-Unassigned Devices-NFS Shares. I.e the mount button worked and was not grey. Now even if create a new share, select 'Load Servers' (which returns the FQDN above) and 'Load Shares', the button is 'MOUNT' button is grey and will unable to interact. If I manually put in the IP address, then the share is mounted and 'MOUNT' button works. Log shows: unassigned.devices: Remote server 'SPIKE.HOME.SOMENAME.ORG' cannot be found on the LAN. If I manually create the mountpoint and run the command, the mount button in the UI will become available but the mount point will no longer be a link: /sbin/mount -t 'nfs' -o rw,soft,noac,noatime,nodiratime,retrans=4,timeo=300 'SPIKE.HOME.SOMENAME.ORG:/docker' '/mnt/remotes/SPIKE.HOME.SOMENAME.ORG_docker'
-
NFS Rules length limit / truncation
years later and this is still a problem (@uproden 's workaround worked for me), how do we get this fixed?
-
Backup to INTERNAL drive with email notifications (and versioning?)?
If anyone is searching for this. I ended up going with Duplicacy. (First was Duplicati but the backups were slow and getting stuck.) You just need to use the unassigned devices plugin. Then one of your internal drives can be an extra backup location and use Duplicacy/Duplicati or something else to sync there. But as alluded to above, this will help protect against your array failure, but not fires, theft, lighting strikes, surges, or the controller/PS frying your backup drive (though Duplicacy versioning should mitigate ransomware). But everyone should have multiple backups (one off-site) and not rely on one backup anyways. I personally also backup to Crashplan (and plan on storing a drive at my brothers house) so it wasn't worth it to me to backup to USB for the very rare chance that controller card/PS fries 2 array drives and backup drive (and not USB drive), bother with expensive tapes or removable media, or build an entire other machine just for additional backup. Also: for email notifications duplicati hooks in well with healthchecks.io and you can use a emailToSMS gateway (check your email-to-number with your provider) to get an txt right to your phone for free whenever a backup doesn't complete
-
Backup to INTERNAL drive with email notifications (and versioning?)?
I'm willing to take that risk as I have other backups (including cloud).
-
Backup to INTERNAL drive with email notifications (and versioning?)?
New to Unraid, I see the backup options are as varied as the sources and targets (makes sense). I've searched for this, but cant find it, even though I'm sure it's somewhere. I'd appreciate it someone can point me in the right direction (guide or just just the right words). Can I If I want to backup my array to an INTERNAL drive. Say 4 + 4 TB Array to 8TB drive. Would it have anything to do with adding this drive as a pool device? An unassigned device? Is there a plugin solution that does email notifications on fail (IMPORTANT) and any chance of versioning (less important), or do I need to try and write something of my own using rsync? Thanks for taking your time to help out newbies like me!
-
New Build comments?
So looks like: Asus PRIME Z590-A ATX LGA1200 Intel Core i5-10400 would be a newer generation (lga1200) at about the same price (or slightly more). Thoughts? Or I'm happy to entertain other options (with so many choices, I'm doing a bit of rolling of dice on pcpartpicker).
-
New Build comments?
Thanks for the reply! Thanks about the CPU. So I notice there are at least two types of NAS builders 1) the buy-the-5-year-old-hardware-types (oft hang out in serverbuilds) 2) the buy-the-newest-stuff types and (everywhere in between). I don't mind being more of Type 1 and sacrifice fade-out if i can swap to i7 in 3 years (if needed) and save some bucks. That said, I'm CERTAIN I also don't fully understand my options. It looks like I can get a new MB and used cpu off ebay for around 300-350 USD. Is there a combo you would recommend that is around the same price (that has my slot requirements) that I should consider?
-
New Build comments?
Hello and thanks for reading. Newbie here, if you have time, care to comment on new build and changes you might make? Past: Freenas (got sick of samba permissions); present: Synology (it's fine); Future: Unraid (easy to grow, less wear on drives, more control) Current: NAS (Synology) 4x 3TB, 2 parity drives ~6T storage total, Essential: 3.5T used Nonessential: (media etc) 4TB stored on 8TB external USB Local backup of essential to external USB 4T Cloud backup everything to Crashplan Backups and storage locations are a bit of a mess App server hosting VM/Docker, inc Plex (ubuntu) 10Gb switch, app server, one music/video production workstation all with SFP+ 10gbe Requirements: Already have Ubuntu application server severing VMs and Docker (inc plex) No high-resource need for VM/docker on unraid at this time, but may want to use in future to grow experiment over time and no desire to combine file server and app server at this time Same space or more without too much more money (based on drive cost) 10gbe iGPU Looking to future: Support Win 11 Slots for HBA, GPU (and 10gbe above) Desired: Would like all machines inc App Server's Docker and VMs and wife's Music/video production to read/write directly from file server 10gbe >300Mb/s for read/write (though I know 10gbe is tricky with Unraid) Trying to keep cost down BUILD: CPU: Intel Core i5-9600KF - Would an i3 be fine here (knowing I have an upgrade path in the future? MB: Gigabyte C246-WU4 - has the slots, m.2, and ecc support for possible future expansion HBA: none (will use board SATA headers, but want to allow for future expansion) 10gbe: 10Gb PCI-E NIC Network Card, Single SFP+ Port, with Intel 82599EN Drives: 2 x 3TB WD Red as storage drives - do I need Pro? 1 X 4TB WD Red as parity 1 x Samsung 980 500GB NVME as Cache PSU: Corsair RM 650W Modular Cooling?? So I think I'm in the ballpark (or at least the parking lot), but with so many CPU/MB/Drive choices I'm sure I'm not on field, I'd love to hear anyone else's insight as I'm not in this space as much as many are. Thanks so much for those that give your time to help out muppets like me! It's REALLY appreciated!
53kl2s
Members
-
Joined
-
Last visited