Beiträge von elastic

    Emergency head retracts makes a noise. Generally, this occurs during a dirty shut down when power is suddenly removed from a spinning hard drive. To prevent a head crash, drives park their heads while the platters are still spinning.

    Your motherboard is "aware" of a soft shutdown and sends the appropriate command to SATA ports to park hard drive heads gracefully. I'm guessing that your add-on HBA card is not doing the same thing. While there is a distinctive "click", I wouldn't worry about it. Your drives are doing what they're designed to do.


    I forgot to post here in case anyone else was wondering, I did find a solution to this via google.


    According to another page I found, these emergency head retracts are quite harsh on the HDD and will ultimately reduce its lifespan. I added this little script to startup that seems to allow the discs to gracefully power off again as they used to before I started using the HBA.

    Bash
    #!/bin/bash
    
    for i in /sys/class/scsi_disk/*/manage_start_stop; do echo 1 > $i; done

    If anyone sees a problem with this script, please feel free to let me know and why it's a problem (so I can learn :))

    The card you linked to on Amazon will be fine. But it's twice the price of the same card on ebay.


    You will need two forward breakout cables to connect the card to the drives. Many of the cards on ebay come with these cables. But you can get them separately too. Use the shortest cables that will reach. Most common are 1 meter (3.3ft) but 1.6ft are also available.


    https://www.amazon.com/gp/prod…h_asin_title?ie=UTF8&th=1

    Thanks. I'm already back up and running :) I did get the shorter cables though

    That leaves a (forever) unanswered question:
    Did the CPU fry because the thermal paste was no longer conducting heat to the sink?
    (OR)
    Did the thermal paste get crusty because the CPU malfunctioned and got to hot?

    In either case, use the good stuff (Artic Silver or something similar) on the next CPU.

    Good question that I sadly think don't ever will be answered. I had thermal throttling enabled on the motherboard, so I was under the impression getting to hot was not possible (for too long anyway) and the system would just run at slower speeds. I had used MX-4 and installed it in mid 2017, and the system only had about 35k hours on it before this failure.


    I've since purchased thermal grizzly as my replacement as I'm not sure this MX-4 paste is still good in the tube. Also after seeing what it turned into, I'm not sure I even wanted to use it again even if it wasn't expired.

    Emergency head retracts makes a noise. Generally, this occurs during a dirty shut down when power is suddenly removed from a spinning hard drive. To prevent a head crash, drives park their heads while the platters are still spinning.

    Your motherboard is "aware" of a soft shutdown and sends the appropriate command to SATA ports to park hard drive heads gracefully. I'm guessing that your add-on HBA card is not doing the same thing. While there is a distinctive "click", I wouldn't worry about it. Your drives are doing what they're designed to do.

    Thanks for the reply. I know I shouldn't worry, but I tend to worry with such things lol. Is there any sort of setting I can change to help with this, whether it be in the motherboard itself or the hba card?

    You are welcome :thumbup:

    I got everything up and running a few days ago. Silly question for you though..


    I've noticed now when I turn my machine off, the hard drives make a noise when shutting off. They never used to do that (I have 8 total) when they were all connected directly to my old mother board. Ever heard of something like this? Everything else works flawlessly btw.

    I only bought one and I chose one of the cheap ones. I don't like ebay and it's what was easiest for me to buy. So far it works without problems after two years. In case it helps you, this is it. https://www.amazon.es/Syba-SI-…rjeta-Verde/dp/B00AZ9T3OU

    In the forum they usually recommend LSI.

    Thanks a lot for the link. So something like this should be ok?


    https://www.amazon.com/Original-9211-8i-FreeNAS-unRAID-6Gbps/dp/B082HKNCVD/


    I noticed your card has the actual SATA ports, but this one has two ports that pigtail into 8. Same thing I'm guessing?

    It is the firmware that the card has installed. If the card already has it installed, then it's done. If the card does not have it installed with a version with TI mode you will have to install the firmware yourself,... if it can be done on the card you buy.

    Thanks, I'll just look to purchase one already flashed with the IT firmware. Any decent brands you can recommend since I'm completely new to HBA cards?

    So my motherboard recently died on me. The motherboard I was using had 8 SATA ports, which I plugged my 8 HDD's to use for my ZFS pool. It's looking like I won't be able to replace this motherboard with something similar, so I'm going to have to use an HBA. Will my ZFS pool be lost once I'm up and running again or will I be able to import it?

    I've been having random kernal issues for a couple of months now. Random things not working, etc. Now today, I can't boot. I admit I tried to downgrade to 5.15 via prox mox before this happened, but now I can't even boot and I'm not sure how to proceed without losing my OS data.


    UUd4Bc1.jpg

    No dice. When I run upsc ups@my.server in SSH, I get this.


    Init SSL without certificate database


    However, it shows me all the stats from the UPS so it does seem like it's connected? I just wish it would stop sending me so many email alerts saying connected disconnected.

    So I have misunderstood. I thought you already solved it.

    Just try creating an SSL certificate manually in OMV to make it available to NUT, it won't hurt you at all.

    In the GUI System>Certificates>SSL click on Create Fill the fields with whatever you want, it doesn't matter what you write. Just establish a 10-year expiration period and write a fictitious email address (with email format or it will not be accepted).

    If it works with that I'll keep that citation in the guide.

    Created the certificate. Do I need to import it anywhere?

    Quote from chente

    You may run into problems if you do not configure SSL certificates. See here RE: Communications with the UPS ups are lost


    I don't SSL at all on my NAS. The plugin was working fine up until a couple of months ago. I have an APC UPS and things worked flawlessly, even the 30 second countdown upon power loss. I'm not really sure how to fix this issue as I really have zero desire to install SSL certs on my machine.

    So apparently this has to do with SSL certs, or in my case a lack of since I do not use SSL certificates on my server?


    Description: status failed (1) -- Init SSL without certificate database


    Someone else posted...


    "Your problem is due to the new NSS backend in nut, which was added in 2.7.1 (Changes from 2.6.5 to 2.7.1). The client tries first to contact tcp/3943 with a STARTTLS command and gets ERR FEATURE-NOT-CONFIGURED from upsd, so it shows that warning."