Jump to content

bubiman

Members
  • Posts

    32
  • Joined

  • Last visited

Posts posted by bubiman

  1. On 6/25/2023 at 10:13 AM, SidM said:


    Servus, 

     

    um die Größe zu verändern musst du erst den Dockerdienst (nicht nur die Container) stoppen. 
    Danach erscheint im selben Menü die Größe die du dann verändern kannst. 

    Danke für deine Hilfe. Hab jetzt die Grenze auf 30GB erhöht. Mittlerweile habe ich auch nicht mehr die ioBroker im Verdacht, der scheint nur "Opfer" eines anderen Dockers geworden zu sein

  2. On 6/23/2023 at 12:34 PM, Kilrah said:

    Then iobroker is probably occasionally crashing and filling up the entire RAM and getting killed. You could add "--restart=unless-stopped" to its extra parameters, and install netdata. After it happens again watch netdata and see if there was a spike in RAM usage in this or another container.

    Thanks, I have installed netdata and will look into it once it crashed again

  3. 15 hours ago, warp760 said:

    Der RAM interessiert niemanden... Ich rede vom Speicherbedarf der Docker. Ist im Standard auf 20 begrenzt 

    Ok da hab ich das dann falsch verstanden, ich hoffe das hier ist die gesuchte Info.
     Demnach ist es noch nicht voll aber fast. Wie und wo kann ich denn diese 20GB vergrößern?

    image.png

  4. 17 hours ago, warp760 said:

    Sind die 20GB von Docker vollgelaufen? 

    Ne, der ioBroker braucht zZ 2.5gb RAM und alle Docker zusammen überschlagen ca. 11gb RAM

     

    Hm oder ich verstehe die Anzeige im Dashboard falsch. Laut der Grafik

    image.png.1359e54c48028626d70b1803a95c7ebc.png

    Ist ja kaum noch RAM frei

  5. Hallo,

    ich habe jetzt zwei mal einen "Out of memory" Fehler durch die ioBroker bekommen:

    Jun 20 17:30:10 Tower kernel: Out of memory: Killed process 18001 (iobroker.js-con) total-vm:32983268kB, anon-rss:21616508kB, file-rss:0kB, shmem-rss:0kB, UID:99 pgtables:49380kB oom_score_adj:0

     

    Eigentlich sollte Unraid mit 32gb RAM genug Speicher haben. Der ioBroker Docker scheint auch bzgl RAM nicht begrenzt zu sein. Wenn ich jetzt in Unraid schaue sind nur ca. 30% der 32gb in Benutzung.

     

    Hat jemand eine Idee was ich versuchen könnte um den Fehler genauer einzugrenzen oder zu beheben?

  6. On 6/12/2023 at 9:13 AM, Kilrah said:

    Note that if a particular container is set to have a max memory allocation precisely to avoid affecting the rest of the server if it misbehaves and exceeds it it'll also be logged as such even if the server itself never got close to running out of RAM.

     Is there anyway to see wich container causes the problem? Yesterday I got the same error again "Out Of Memory errors detected on your server".

     

    The docker tab suggest, that there is no max memory allocation set:

    image.thumb.png.db5d9b49805087abeafbbde7d2dcd855.png

    Every docker seems to have the whole 32gb available

  7. Hi,

    I have to admit, I am not checking my unraid server on a daily basis. Today I saw a red error from the "FIX COMMON PROBLEMS" Plugins:

     

    Out Of Memory errors detected on your server

     

    I have not noticed any malfunction from the server and do not know what to do except the recommended reboot of the server from this post: 

     Should I check or do somthing else before rebooting?

     

     

  8. Hi, I am having problems mounting an encrypted USB drive.

    I activated the destructive mode, deleted all partitions, formated the drive with xfs-encrypted ... that all went well:

    Jan  5 20:07:00 Tower unassigned.devices: Removing partition '1' from disk '/dev/sdb'.
    Jan  5 20:07:55 Tower unassigned.devices: Format device '/dev/sdb'.
    Jan  5 20:07:55 Tower unassigned.devices: Device '/dev/sdb' block size: 5860533168.
    Jan  5 20:07:55 Tower unassigned.devices: Clearing partition table of disk '/dev/sdb'.
    Jan  5 20:07:55 Tower unassigned.devices: Clear partition result: 1+0 records in 1+0 records out 2097152 bytes (2.1 MB, 2.0 MiB) copied, 0.013233 s, 158 MB/s
    Jan  5 20:07:57 Tower unassigned.devices: Reloading disk '/dev/sdb' partition table.
    Jan  5 20:07:57 Tower unassigned.devices: Reload partition table result: /dev/sdb:  re-reading partition table
    Jan  5 20:07:57 Tower unassigned.devices: Creating Unraid compatible gpt partition on disk '/dev/sdb'.
    Jan  5 20:08:00 Tower kernel: sdb: sdb1
    Jan  5 20:08:00 Tower unassigned.devices: Create gpt partition table result: Creating new GPT entries in memory. The operation has completed successfully. 
    Jan  5 20:08:02 Tower unassigned.devices: Reloading disk '/dev/sdb' partition table.
    Jan  5 20:08:02 Tower kernel: sdb: sdb1
    Jan  5 20:08:02 Tower unassigned.devices: Reload partition table result: /dev/sdb:  re-reading partition table
    Jan  5 20:08:02 Tower unassigned.devices: Formatting disk '/dev/sdb' with 'xfs-encrypted' filesystem.
    Jan  5 20:08:11 Tower unassigned.devices: Format drive command: /sbin/mkfs.xfs -f '/dev/mapper/format_sdb' 2>&1
    Jan  5 20:08:26 Tower unassigned.devices: Format disk '/dev/sdb' with 'xfs-encrypted' filesystem: meta-data=/dev/mapper/format_sdb isize=512    agcount=4, agsize=183140635 blks          =                       sectsz=4096  attr=2, projid32bit=1          =                       crc=1        finobt=1, sparse=1, rmapbt=0          =                       reflink=1    bigtime=1 inobtcount=1 data     =                       bsize=4096   blocks=732562537, imaxpct=5          =                       sunit=0      swidth=0 blks naming   =version 2              bsize=4096   ascii-ci=0, ftype=1 log      =internal log           bsize=4096   blocks=357696, version=2          =                       sectsz=4096  sunit=1 blks, lazy-count=1 realtime =none                   extsz=4096   blocks=0, rtextents=0
    Jan  5 20:08:29 Tower unassigned.devices: Reloading disk '/dev/sdb' partition table.

     

    After that I set the password for the drive in the UNASSIGNED DEVICES settings:

    Jan  5 20:09:03 Tower  ool www[25856]: /usr/local/emhttp/plugins/unassigned.devices/scripts/rc.settings 'set_pass'
    

     

    and hit the mount button on the main tab and it is not working:

    Jan  5 20:27:22 Tower unassigned.devices: Adding partition 'UnraidBackup1'...
    Jan  5 20:27:22 Tower unassigned.devices: Mounting partition 'UnraidBackup1' at mountpoint '/mnt/disks/UnraidBackup1'...
    Jan  5 20:27:22 Tower unassigned.devices: Using disk password to open the 'crypto_LUKS' device.
    Jan  5 20:27:25 Tower unassigned.devices: luksOpen result: No key available with this passphrase. 
    Jan  5 20:27:25 Tower unassigned.devices: Partition 'IB-1122- U3' cannot be mounted.

     

    Does anyone have an idea what I am doing worng?

    Thanks for your help

  9. 1 hour ago, inf said:

    This issue caused me some headache as well...

     

    Turns out the Intel C3000 Atom platform has support for something called Intel QuickAssist Technology (Intel QAT, https://www.intel.com/content/www/us/en/architecture-and-technology/intel-quick-assist-technology-overview.html). The kernel driver for this feature is what is causing the issue with LUKS/dm-crypt.

     

    The QAT kernel driver takes precedence when both AES-NI and QAT is activated so the solution is to either blacklist the kernel module or disabling the QAT feature in BIOS. I have a Super Micro board and the setting for QAT was found under the south bridge configuration (IQAT).

     

    As a bonus the performance with AES-NI compared to QAT was twice as fast (according to cryptsetup benchmark)

    Thanks for your follow up. I had given up on using encryption with unraid but I will look into turning of QAT through BIOS  and try it again.

     

    bubiman

  10. To give an update (I hate to find a topic describing a problem but there is no final post):

     

    The old USB stick is unsavable. No matter what I have tried, it remaind "write protected".

    I prepared a knew USB stick with the Unraid USB creator and after that was finished I copied my backup over the files. That was not working, my PC did not want to boot from that stick.

     

    So I cleaned everything from the stick and just copied my backup onto it and used the make-boot.bat.

    That did work, my tower is ruing again. I had to got a new key for the stick.

    • Like 3
  11. I have a backup through the web frontend of unraid. What else should I backup to be on the safe side?

     

    27 minutes ago, itimpi said:

    Something like Kingston Data Traveler (ones with metal case) are a good choice.    Note that USB2 versions are recommended as being more reliable (as well as cheaper) than USB3 variants.

     That is the one I have ordered.

  12. Hello,

    today I noticed that my USB stick seams to be failing.

    I have read some other threads regarding this error. I have already ordered some new USB sticks because the stick I am using rigth now (it is an Sandisk Ultra Fit, one of these tiny ones) is not recommended for unraid according to some entries here.

     

    Should I even try repairing (I would follow these instructions 

    ) the stick or should I directly migrate to a new USB stick? I do not want to waste my "one new usb stick every 12 month" but the current stick just lasted 8 month. Is there a way to check if the stick is realy broken or if just something went wrong?

     

    Thanks for your help

    Anmerkung 2020-07-11 222015.png

×
×
  • Create New...