This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

RPM error on new SOPHOS UTM 9 install

Recently I started to get a number of issues on my SOPHOS 9 box. Since I'd been intending to replace the spinning rust with an SSD, I gave in and did so (checking the old disk confirmed it was generating a rapidly rising number of SMART errors). However, so far I've been unable to re-install. Every time I've tried to date, towards the end of Stage 5 it comes up with an RPM error. I've tried the install with three different USB sticks, and two different ISO writers, including RUFUS. I've tried with 9.502-4.1, 9.414-2.1 and 9.358-3.1. All the hardware is the same as when it was working, apart from the new SSD. I have the install.zip support file if it would mean anything to anyone :-). Might anyone have any thoughts?



This thread was automatically locked due to age.
Parents
  • After selecting the disk where Sophos will be installed, go to the console mode (ALT-F2), then enter:

    mount /dev/sdb1 /install

    Then switch back using ALT-F1 and continue. See if that helps.

  • @exp3371

    My thanks indeed for the reply :-). I was,in fact, already remounting sdb1, but it is indeed something that gets missed a lot.

     

    As a more general update, I spent some time on it this weekend. I went back to a clean slate and started from scratch. Results:

    1: Three releases of the ISO image all generated RPM errors.

    2: As a black box test, I downloaded a MINT ISO and tried to install that. The install failed on the first file write.

    3: Assuming I had either a hard disk issue or a controller issue, I tried the SOPHOS install on three different, working spare hard disks I had available. All white screened with what generally appeared to be controller related errors almost immediately the install started.

    All the test so far have been with the same RAM stick - I haven;t tried changing that. There's (to my limited wit) an outside chance it's a RAM fault, but I'm betting it's a controller issue. Unfortunately the motherboard in this particular case is an embedded board. The board has one PCI slot which, at the moment, I have filled with a dual Intel NIC (to avoid using the onboard Realtek NICs). So I can't easily put in a new controller to test it. Plus, if there really are controller issues, I'm not betting against there being other motherboard issues.

    So my courses of action appear to be:

    1: Do nothing for now. I have a SOPHOS running as a VM - this one was to be the second of a failover cluster pair.

    2: Get a new embedded motherboard for the micro case and build the second SOPHOS.

    3: Give in to temptation, and get a ZOTAC C series instead - I have one I use for audio recording work, and the total silence, small size and low power use are itchingly attractive... (blush). 

    Given the tests so far, I'd be interested in any views that support or reject the controller conclusion - and any thoughts on a ZOTAC C as a SOPHOS host :-).

  • Had a similar experience a while back with RPM errors. It turns out it was the USB drive and the ISO to USB software that I used weren't good. Changed to Sandisk USB drive and used Rufus to build the USB boot disk and was fine after that.

    I don't have any experience with the Zotac C series. I took a chance on a Protectli 4 Port J1900 – 8GB RAM / 120GB mSATA. So far so good after a few months.

Reply
  • Had a similar experience a while back with RPM errors. It turns out it was the USB drive and the ISO to USB software that I used weren't good. Changed to Sandisk USB drive and used Rufus to build the USB boot disk and was fine after that.

    I don't have any experience with the Zotac C series. I took a chance on a Protectli 4 Port J1900 – 8GB RAM / 120GB mSATA. So far so good after a few months.

Children
No Data