niietzshe

Members
  • Posts

    71
  • Joined

  • Last visited

Posts posted by niietzshe

  1. Thanks all. I mean I’ve had a good run of at least 8 years without a failure, so it’s a good thing I’ve never had to deal with this I guess.

    I’ve tried the SSd in multiple ports just to make sure and it’s certainly knackered. I’ve bought a new SSd and will be shopping for a ups and little monitor to go in the cupboard.

    Thanks for the advice
    Christian


    Sent from my iPhone using Tapatalk

  2. Thanks Michael, I quickly disconnected the ssd before I ran out and it booted straight into unraid, so have ordered a new SSD.

    Now here’s a question:
    Seeing as unraid is on the usb flash drive, why didn’t this boot into unraid and then I would be able to see that it’s a busted drive? I’m thinking of getting a second SSd for redundancy, but this wouldn’t really help me as it would still stop unraid from booting to discover that I need to replace the disk. As it’s totally headless I don’t see the bios boot until I move the server into my office and hook it up to a monitor.


    Sent from my iPhone using Tapatalk

  3. Ahhh.
    So the Kingston drive is just my stash drive or whatever it’s called. The SSD. It might hold my VM, which will be a pain, but I can sort that out.

    I’ll remove it when I get back and see if that helps it. I have turned off the smart detection but that didn’t resolve the issue.

    Thanks for the tips. I’ll report back. Time to get a UPS!

    Christian


    Sent from my iPhone using Tapatalk

  4. Hey all,

    my servers been up and running for like 8 years with no real problems, probably only restarted once or twice last year, uptime has been great.

     

    The other day we had a power cut and now i cant get it to boot into the UnRaid flash drive automatically. when it fails i can force it to start and everything is fine, but ive had to move the server away from the cupboard to do that and get a monitor connected.

     

    Any tips on how to debug this? Ive tried to force it to the sandisc in bios, but the hba always takes over and tries to boot from the raid drives.

     

    thanks

    christiab

    03467178-8C4C-4ED0-9E40-22C3906011B1.jpeg

    98C78075-7C75-42DD-A337-10141A2B3DF2.jpeg

    B194716D-66C0-43DD-AF4F-84E190D1A194.jpeg

    1AE38769-AA43-4542-9452-B64F88C3A3B5.jpeg

  5. Thanks. I honestly can’t remember, but I’m sure I connected to the header on the motherboard. I’ve checked ambient temp, it’s 25c a few inches in front of the case.

    I’ll check my fans and your suggestions, although it’s a pain to get to bios as I have to rip it all out and connect it to my tv. But I’ll have a go.

    I was going to upgrade the case and move to Matt for a graphics card, but apart from plex hardware decoding, I no longer have much of a need for it (I was going to do some 3d work at home, but now I can just vpn into the studio thanks to covid). But a bigger/better case would have given me better airflow.


    Sent from my iPhone using Tapatalk

  6. Hi all. This is long overdue. My Lian-Li build has had fans on constantly for years now. I’ve always meant to look into why, but thought it was probably just it’s environment (in a large cupboard under the stairs). But we recently opened up said cupboard to change its configuration and haven’t noticed any change in temps or noise.

     

    Build:

    LianLi q25b

    ST30SF PSU

    2 Gelid silent pwm fans

    ASUS H81I-Plus motherboard

    8gb ram

    120gb ssd

    4x 3TB drives (WD greens/reds)

    i5 4670k

     

    Applications:

    usual plex/nzb dockers

    home assistant vm

     

    Average drive temps are 45C

    Cpu is 48C at 6% load

    Fans are 2700 and 860 rpms

     

    Im open to suggestions. It’s just something I know very little about. If it’s a software tweak, or a hardware change, just throw ideas at me.

     

    When we build up the under stairs again I’m going to have a little area at the top. So I might install in and out vents under there to bring the temps down if I can.

     

    Thanks for your time.

  7. Well after a night of parity checks I normally have an endless list of notifications telling me all disks have overheated, normally well into 50’s or even 60s sometimes.

     

    This morning was a lot better!

    96% done and fans running 100% still.

    Im still interested in making this better if anyone has any tips/thoughts/comments.

     

    B7BFED86-BECF-4D3F-877E-797BC8B8BA04.jpeg

  8. Hey all, sorry I just do not know anything about pwm fans so have to ask some pretty basic questions as I've got hot running drives. Or had, they might be more stable now.

     

    My motherboard is a Asus H81I-PLUS, I have upgraded the fans to pwm fans and installed the dynamix temp and fan control plugins. They seem to be working. I have an airflow field on my dashboard now, one running at 1521RPM, the other at 1236RPM. I then have a cpu fan, but that's not registering anywhere.

     

    Processors temperature (pentium G3420) is at 45C.

    My drives (4 spinners, 1 SSD) are running at an average of 37C, SSD at 26C.

    I think previously my fans just weren't working, so those drives would hit 60C easily under load.

     

    So I just want to double check I have everything setup correctly.

    Under Fan Auto Control - PWM controller, I have selected nct6779-pwm1, but there's also pwm2, 4 and 5.

    The fan it's found is '/sys/devices/platform/nct6775.656/hwmon/hwmon2/fan1_input' but why only one fan?

     

    Under the temp plugin, the available drivers are 'coretemp nct6775'.

    Processor temp = coretemp - CPU Temp - 44C

    Mainboard temp = acpitz - MB Temp - 27.8C

    Array fan speed = nct6799 - Array Fan - 1508 RPM

     

    Because these have different options and I'm slightly confused why it's only dealing with one fan, I'm here asking for guidance. Is what I've done correct? Could I change things to make it better?

    The case is quite small, a Lian-Li q25B m-itx case.

     

    Additional info:

    MB   = Asus H81M-PLUS

    CPU = Intel Pentium G3420

    PSU = SilverStone SST-ST30SF

    Fan1 = Gelid Solutions Silent 14 PWM

    Fan2 = Gelid FN-PX12-15 Silent 

     

    I've included a diagnostic and some screenshots of drop down choices, just incase that's helpful.

    Thanks for any advice.

    Christian

    marvin-diagnostics-20190902-2305.zip

     

    Screen Shot 2019-09-02 at 23.41.45.png

     

    Screen Shot 2019-09-02 at 23.42.23.png

     

    Screen Shot 2019-09-02 at 23.42.36.png

     

    Screen Shot 2019-09-02 at 23.42.31.png

  9. Hey all, I've had a consistent problem for the past 5+ years with UnRaid taking down the cache drive when a parity check happens (monthly). For this reason I've been running an NUC for Home Automation dockers as this needs to stay up, or at least be back up by the morning. I'd like to put an i7 in and return everything back to UnRaid, but I'll have to solve this issue. I'm currently running 6.7.2 but it's been happening for a long time.

     

    IMG_7951.jpeg.c049b596fefecf7a536daf98b4e27a43.jpeg

     

    I've attached my syslog, it would be great to get this resolved.

    Thanks

    Christian

    syslog.txt

  10. Really?

    I've been looking at building a new machine to upgrade a mitx intel pentium build. Should I be avoiding Ryzen still?

    I was looking at the same motherboard but a 1600 cpu for esxi with unraid as guest along with some other VMs and dockers (most intensive being plex of course).

  11. If parity checks do cause a problem then you can schedule them to only be run at times when the server would normally be idle.


    I normally have to reboot after a parity check (3am). It takes dockers down...

    There’s enough power in the NUC for what I need, it stays at 40%. I was just wondering if I should start a new path.


    Sent from my iPhone using Tapatalk
  12. That’s true,
    I’m currently monitoring Plex with tautuli through home assistant to see what’s going on.

    To be fair the easiest thing to do is replace my router with an R7800, an Apple TV upstairs to replace the nexus player and hardwire it. Then I probably wouldn’t be transcoding at all.


    Sent from my iPhone using Tapatalk

  13. Hey all, Bare with me while I slightly ramble.

     

    I've been running UnRaid on a miniITX board in a LianLi Q25b case for about 5 years. It's a Pentium3.2ghz build with 8GB DDR3 ram, 3x3TB data, 1x3TB parity and a 120GB SSD for cache. I run a number of dockers, mainly Plex is the biggie, no VMs. I also run an IntelNUC Celeron on Ubuntu for home automation, this is also up 24/7 and runs lots of Dockers like Home-Assistant, Node-REd, Mosquitto, MariaDB etc.

     

    The reason I moved all that stuff to an NUC instead of keeping it in UnRaid was because of Parity checks taking it down which wasn't a good WAF.

    I'm pretty happy with the setup. I do see plex spiking the CPU, but I mainly just don't like having two places to maintain dockers and the NUC won't run Plex very well. I also need to Update my Router (a Linksys e4200....old!).

     

    So I started thinking I could upgrade my UnRaid server, run Proxmox, Virtualize Unraid as a guest and also run my dockers in another VM and even PFSense in another. This would require me to rebuild, get a Fractal Node 8, new MB, new CPU, probably another 8GB ram and a new NIC card. I'm hoping that would be all.

     

    What do you guys/gals think of this approach? Would you do it? Is it a big no-no bringing it all together into one. Any alternatives or suggestions? I'm thinking of building it slowly over a few months, so could look out for eBay deals, I'm not after blazing, just enough+.

     

    Thanks for any thoughts on the matter.

    Christian