-
Posts
378 -
Joined
-
Last visited
Content Type
Profiles
Forums
Downloads
Store
Gallery
Bug Reports
Documentation
Landing
Posts posted by Mainfrezzer
-
-
Bin mal neugierig. Mit ist aufgefallen dass seit meinem Update von 6.11.5 auf 6.12.4 mein Messgerät nen ~1 Watt höheren verbrauch meldet. Kann das jemand bestätigen? Hab das gerade nochmal nachgeguckt und bei 6.11.5 komm ich auf 14.4 - 14.9 Wattbereich wenn nix außer 1 Docker Container läuft und bei 6.12.4 ist es 15.3-16.1 Watt.
Bei "Semi-Last" war ich bei knapp unter 20 Watt unter 6.12.4 und bei 6.11.5 ist es 17.3-17.7 Watt -
11 minutes ago, swamiforlife said:
i dont want other users on the network to probe or see that this server exists.
Thats practically impossible on the same network. As @JonathanM said, thats a good way. Otherwise iptables would be a friend.
-
You have to have something, a usb thumbdrive for example, in Disk 1 of the Array. Currently you're unable to start "Unraid" without a single device in the "Unraid Array"
-
Unraid has been up for a couple of days and absolutely nothing popped up on my system. neat.🥳
-
1
-
-
Make sure to let your friend know aswell
-
1
-
-
5 hours ago, MSMusixs said:
Das liegt daran ->
5 hours ago, MSMusixs said:Wenn man unter Internet->Zugrangsart-DNS Server den Eintrag setzt, dann ist die Fritzbox immer noch der "Hauptdns". Die Fritzbox fragt dann für alle Clients den DNS Server an.
Man müsste es über "Heimnetz - Netzwerk -> Netzwerkeinstellungen" -> Runter scrollen und weitere Einstellungen links unten klicken -> IPv4-Einstellungen und auch IPVv6 Einstellungen eintragen.
Dort sind für die jeweiligen Protokolle die DHCP Einstellungen. Dort müsste man die ips für den gewünschten DNS Server eintragen. Dann bekommen die Clients die Infos beim nächsten verbinden mitgeteilt.
In dem 1. Bild wurde dort die GUA (die 2a02:8071: Adresse) eingetragen, das ist ein Problem. Da muss die fd00 Adresse oder fe80 rein.
5 hours ago, MSMusixs said:Auch hat in Docker/Unraid jeder Container immer noch nur die IPV4 Adressen (in der fritzbox sind aber IPV4 und 6)
laut docker manual muss im daemon + custom network ipv6 aktiviert werden.
Unter anderem soll in /etc/docker/daemon.json "iptabeles" aktiviert werden
When using IPv6, we recommend that you enable the experimental ip6tables Parameter in the daemon configuration.
Der Ordner /etc/docker/ ist auch vorhanden, nur die daemon.json ist nicht drin..
Wo packt Unraid die hin???
Ich bin mir da nicht ganz sicher worums hier geht. Die Anzeige der Docker Seite? Da stehen nur die ipv4 Adressen. Dass das ganze sonst richtig funktioniert ist ja schon erledigt. Deine Container bekommen ja ipv6 Adressen.
5 hours ago, MSMusixs said:Ich würde die native ipv6-Anbindung verwenden, wenn man ipv6 benutzt. Das ist aber nur relevant für Internet-Verkehr, intern ist das irrelevant.
Die eine IP Adresse, 127.0.0.1, das ist ne spezielle IP. Das ist localhost. Die Fritzbox kann da nur sich selbst ansprechen mit. Nur so nebenbei. -
Der Fehler ist gefunden.
Wenn man macvlan nutzen möchte und auf 6.12.4 ist, muss man die bridge deaktivieren. Steht auch hier https://docs.unraid.net/unraid-os/release-notes/6.12.4/ -
28 minutes ago, Deadeye said:
I am already at version 6.12.4!
You didnt read the notes that state to disable the bridge if you use macvlan.
Edit: ive had to edit this so much cause i opened the wrong log file first, but yeah, disable the bridge/bridges since you do have br0 and br1. -
the auto tune is too aggresive and turns off the NIC when theres absolutely nothing going on on the network and never turns it back on again. Had that happen to me aswell.
# ------------------------------------------------- # Set power-efficient CPU governor # ------------------------------------------------- #/etc/rc.d/rc.cpufreq powersave # ------------------------------------------------- # Enable power-efficient ethernet # ------------------------------------------------- # enable IEEE 802.3az (Energy Efficient Ethernet): Could be incompatible to LACP bonds! #for i in /sys/class/net/eth?; do dev=$(basename $i); [[ $(echo $(ethtool --show-eee $dev 2> /dev/null) | grep -c "Supported EEE link modes: 1") -eq 1 ]] && ethtool --set-eee $dev eee on; done # Disable wake on lan #for i in /sys/class/net/eth?; do ethtool -s $(basename $i) wol d; done # ------------------------------------------------- # powertop tweaks # ------------------------------------------------- # Enable SATA link power management echo med_power_with_dipm | tee /sys/class/scsi_host/host*/link_power_management_policy # Runtime PM for I2C Adapter (i915 gmbus dpb) #echo auto | tee /sys/bus/i2c/devices/i2c-*/device/power/control # Autosuspend for USB device echo auto | tee /sys/bus/usb/devices/*/power/control # Runtime PM for disk echo auto | tee /sys/block/sd*/device/power/control # Runtime PM for PCI devices #echo auto | tee /sys/bus/pci/devices/????:??:??.?/power/control # Runtime PM for ATA devices echo auto | tee /sys/bus/pci/devices/????:??:??.?/ata*/power/control
This is my setting. the cpu govenor is disabled because i franky dont need it and the effient ethernet bit is the problematic one, that also why its disabled. auto tune just turns that one on.
so you can basically remove the auto tune line and replace it with just this
# ------------------------------------------------- # powertop tweaks # ------------------------------------------------- # Enable SATA link power management echo med_power_with_dipm | tee /sys/class/scsi_host/host*/link_power_management_policy # Runtime PM for I2C Adapter (i915 gmbus dpb) #echo auto | tee /sys/bus/i2c/devices/i2c-*/device/power/control # Autosuspend for USB device echo auto | tee /sys/bus/usb/devices/*/power/control # Runtime PM for disk echo auto | tee /sys/block/sd*/device/power/control # Runtime PM for PCI devices #echo auto | tee /sys/bus/pci/devices/????:??:??.?/power/control # Runtime PM for ATA devices echo auto | tee /sys/bus/pci/devices/????:??:??.?/ata*/power/control
For you current problem, not being able to access the server, you can only get it done by hooking up a keyboard to it and running cmds to try to bring the nic back up or shutdown to the server to make the modifications on the usb drive so that never happens again.
Edit:
or also the (Runtime PM for I2C Adapter (i915 gmbus dpb) i cant remember why i turned it off, i know it had a reason but honestly i have no clue why now
Edit #2:
I did see that the PCI Devices is disabled/commented out, thats was on my end the important second line because my NIC is treated as PCI device. But you would have to check how it works in your setup.
So to sum it up, once again
# ------------------------------------------------- # powertop tweaks # ------------------------------------------------- # Enable SATA link power management echo med_power_with_dipm | tee /sys/class/scsi_host/host*/link_power_management_policy # Autosuspend for USB device echo auto | tee /sys/bus/usb/devices/*/power/control # Runtime PM for disk echo auto | tee /sys/block/sd*/device/power/control # Runtime PM for ATA devices echo auto | tee /sys/bus/pci/devices/????:??:??.?/ata*/power/control
Should work out but it depends on what you have as a system.
Thats a link for a bit more information on it
https://forums.unraid.net/topic/98070-reduce-power-consumption-with-powertop/-
1
-
1
-
-
19 minutes ago, Roalkege said:
After an unwanted power outage, I can now remove the container
Life lemons, lemonade
something positive from it at least.
Strolling through the files, i do notice youre running 6.13.3 but run an incompatible go file script.
The 1.4 version of the Ram-log script does not work in 6.12.3
Version 1.6 for 6.12.3
If you decide to upgrade to 6.12.4, the code can be found a bit further down. -
You would need to post the diagnostics because we can only ask the magic crystal ball for this "execution error".
the zfs plugin/ the snapshots in general are not the way. Snapshots are, well, snapshots not the actual data^^ -
Docker container are removed over the Docker page, not the zfs master plugin. Those snapshots are all created by the dockerservice.
-
Do you use powertop with the "--auto-tune" by any chance?
-
Just as a one off random chance shot in the dark.... You're not trying to boot in uefi mode are you? The 730 only works in legacy/bios mode.
-
10 minutes ago, JonathanM said:
My bet is that the battery in the UPS is completely dead, regardless of what the UPS thinks the status is.
I wouldn't bet on that since
1 hour ago, KingLB said:can then power on the server manually and it starts on battery power.
I'm certain it's due to the switch time, especially since the Synology worked fine. Either the ups specsheet is lying or the evga psu is not up to spec
-
mhmm spontaneously i would say
"dd if=/dev/zero of=data.null bs=10G count=100" which would be a 1TB file with lots and lots of null.
"dd if=/dev/zero of=data.null bs=10G count=1000" would be 10TB
Be careful, it creates the file where you currently are in the console.
Edit: You could also download a zipbomb for quick and easy data to fill the empty space with -
Just now, wayner said:
By this you mean that I don't have to actually do a force update? I just have to have it recreate the docker due to a slight change in a parameter? So I could add a new variable called "dummy" with the value "dummy" and I would be good?
Thats even too much efford put in^^
all you need to do is literally this
-
Docker Container nur ne ipv4 zuweisen.
wenn MACVLAN benutzt wird, was ich mal stark vermute bei ner Fritzbox, --mac-address 01:23:A0:B1:C3:D4 eintragen bei den Extra Parametern (sichtbar nur bei Aktivierung der erweiterten Ansicht oben rechts)
Starten
Wenn man lust hat kann man jetzt in der Fritzbox die ULA immer zuweisen lassen, oder auch nicht, egal.
Der Container bekommt dann folgende Adressen zugewiesen.
fe80::0323:a0ff:feb1:c3d4
fd00::0323:a0ff:feb1:c3d4 (wenn man die standard ULA bei der Fritzbox nimmt)
ÖffentlicherPRÄFIX:0323:a0ff:feb1:c3d4
Und was man nun letztendlich für den DHCP Server nimmt ist entweder die fe80 Adresse, nur lokal in einem Netz benutzbar oder die fd00 Adresse, die man für andere Subnetze benutzen könnte (ermöglicht halt auch rDNS für DNS Server, aka statt Adressen stehen dann die clientnamen) -
7 minutes ago, itimpi said:
As I said it is easy to restrict incoming connections to whatever you want. There was no mention of what services those users should be able to access as restricting those may have implications.
Still not on it.
The client that connects to the unraid wireguard server shall have access to the whole of the internet, wherever they are in the world. Everything, except for the network the wireguard server is running on, i.e bogus networks.
Pratically be a tor exit node without the ability to reach the local network of the node.
theres a guide for that but im not sure how well that translate to the unraid wireguard variant
https://gist.github.com/qdm12/4e0e4f9d1a34db9cf63ebb0997827d0d -
6 minutes ago, itimpi said:
I do not understand - are you saying that you want users on your local LAN to access the internet ? As I said it is easy to restrict incoming connections FROM the internet to what you want them to access.
Na not me, the OP.
OP wants to provide a VPN connection for someone outside their lan to use their server as exitnode but does not want the clients to be able to access the lan network the wireguard server is running on. -
1 hour ago, wayner said:
What happens if you don't/can't do a force update to all of your Docker containers? What are the ramifications?
You cant start that docker container
1 hour ago, wayner said:I have one Docker that I don't want to update to a newer version, can I "force update" it back to the same prior version? And I may have a docker that is no longer available.
Use the tag for the specific version of the container (i.e Author/Dockerimage:V1.23)
If that docker image does not exist on the repository anymore, you still have it locally, which will be used unless you delete it OR if you wanna be super save, push it to your own repository as backup.
All you ultimately need to do is a dummy change in the edit fuction, like add a 1 to a random line and delete it, then hit apply and youre done. -
It depends on the power supply, how sensitive they are to losing power.
You probably dont have a spare PSU around to test your server/ups with i assume?
You could look up the hold-up time for your psu and check how fast the switch time is of the ups.
Edit: All i can find for the ups is a "respond time of 4ms" which i doubt because i dont think that a psu from evga is that horrible in terms of hold-up time -
If the PC turns of the second you unplug the cable from the UPS to the Wall, theres something going on with your UPS. Probably not fast enough with switching or something else.
-
43 minutes ago, itimpi said:
It is quite easy to make sure they can only get to your server, or specific IP addresses inside your LAN.
Thats the total opposite^^
Whats sought is Internet only - No Lan access.
Nextclound und der Spin Down
in Deutsch
Posted
Das ist keine Raketenwissenschaft. Die Einstellung ist unter setting -> global share settings