Page 3 of 3 FirstFirst 123
Results 21 to 28 of 28
  1. #21
    Untangle Ninja
    Join Date
    Jan 2011
    Posts
    1,325

    Default

    Quote Originally Posted by sky-knight View Post
    Code:
    Disk /dev/md0 doesn't contain a valid partition table
    I love the final line, obviously if it didn't have a valid partition table it wouldn't have booted! lol
    I get the same thing, so I was just wondering wtf? I don't recall seeing that on the SLES 10 servers I've done with md raid... but maybe I never ran fdisk -l on them.

  2. #22
    Untangler
    Join Date
    May 2008
    Posts
    426

    Default

    That is not a problem. Fdisk jut doesn't know how to read the md, it only handles the partitions that are real. Been that way as long as I can remember on redhat / centos.

    Don

  3. #23
    Master Untangler Big D's Avatar
    Join Date
    Nov 2008
    Posts
    719

    Default

    Software raid does work. Originally the installer didn't hook the mdadm packages correctly and you had to migrate a regular disk to software raid after the install. This was fixed in v.8.0 of the installer.

    Only thing I have seen picky about the installation is which partition you set the root filesystem for think it wants sda1 / sdb1 = md0.

    I am running software raid in lab and at 1 client that is nearby. My brother is using UT install using software raid with asterisk functioning as a IAX gateway back to his servers in the datacenter.

    He has a good 5 or 6 of those out there. Says his clients love it they use it to curb the call centers surfing habits and lock them out of crap their not supposed to be doing. Which leaves more bandwidth for SIP. So far he hasn't encountered any issues.

    Add this to your crontab:
    Code:
    5 0,12     * * *   root         mdadm --monitor --scan -1 --mail=alertservercritical@domain.com
    This should shoot you an email if the md devices go non optimal. Or add a monitoring set to nagios or whatever server management system your using. We have Kaseya but been too busy and lazy to figure out the monitoring sets for linux yet.

    It should be pointed out UT does not support software raid. Should this thread be moved to the hacks forum?
    Last edited by Big D; 06-15-2011 at 08:23 AM.
    The beatings shall continue until morale improves!

  4. #24
    Untangle Ninja
    Join Date
    Jan 2011
    Posts
    1,325

    Default

    Quote Originally Posted by Big D View Post
    Only thing I have seen picky about the installation is which partition you set the root filesystem for think it wants sda1 / sdb1 = md0.
    Aha! That's why my first md raid install didn't work. On the first installation I did with md raid, I put the swap partition on md0, and put root on md1. It's the second installation where I just duplicated the exact partition layout of a default UT install.

    Add this to your crontab:
    Code:
    5 0,12     * * *   root         mdadm --monitor --scan -1 --mail=alertservercritical@domain.com
    This should shoot you an email if the md devices go non optimal.
    I was thinking about wanting that... thanks for spelling out how to do it.

    It should be pointed out UT does not support software raid. Should this thread be moved to the hacks forum?
    It'd be nice if UT would support simple mirror-pair configurations. I fiddled with SMESERVER once, it handles this very cleanly: "oh, we see you have two hard drives, would you like mirroring?". Surely the installation script to duplicate what we've done here couldn't be too difficult, though perhaps Debian won't allow that level of automation.

    But I applaud UT for fixing whatever was broken prior to 8.0 that prevented this from working at all. I did read your guide on the subject (well done sir), glad it doesn't take that level of pain any more.
    Last edited by johnsonx42; 06-15-2011 at 09:01 AM.

  5. #25
    Untangle Ninja sky-knight's Avatar
    Join Date
    Apr 2008
    Location
    Phoenix, AZ
    Posts
    26,228

    Default

    Support for a basic mirror would be nice, however this isn't something that UT should be supporting.

    This type of thing is specifically a value added service that can be provided by the resellers in the field. Jim and I are working on potentially preconfiguring NG series units for a soft mirror configuration. But, those units will only be available to trained resellers.

    If you aren't comfortable with the terminal how are you going to readd the drive after it's "dead"? This isn't point and click RAID! I don't see any issues using it long term, if you have the knowledge to maintain it. Untangle will never really be able to support this because phone support + RAID just doesn't work.

    Oh, and Linux has always been a bit twitchy if /boot isn't the first partition. So unless you want to make your own /boot partition, I always put / in as the first partition.

    So I guess a stockisk partition scheme should be.

    80gb / Primary Bootable
    *gb /data primary
    2xRAMGB SWAP Logical

    You just want the swap at the end of the drive, at least based on default Debian / Ubuntu installer, and everything I've ever heard about the performance of swap space. Of course, if your Untangle is swapping... you need to buy ram!
    Last edited by sky-knight; 06-15-2011 at 09:42 AM.
    Rob Sandling, BS:SWE, MCP
    NexgenAppliances.com
    Phone: 866-794-8879 x201
    Email: support@nexgenappliances.com

  6. #26
    Untangle Ninja
    Join Date
    Jan 2011
    Posts
    1,325

    Default

    ah yes, you're right, I didn't ponder the future considerations of supporting software raid for everyone+dog.

  7. #27
    Untangle Ninja
    Join Date
    Jan 2011
    Posts
    1,325

    Default

    minor point of interest... I re-installed this box tonight before shipping it out, just so it would have a fresh, non-fiddled with installation; I also wanted to use build4, just in case.

    anyhow, when I went into the partitioning tool, I discovered that I hadn't set the Bootable flag on the two raid partitions as was shown in your video, and the system booted fine anyhow. do you suppose the flag is irrelevant when using mdraid, or does it just choose the first partition if none are flagged as bootable?

  8. #28
    Master Untangler Big D's Avatar
    Join Date
    Nov 2008
    Posts
    719

    Default

    oh second hiccup the installer wont install grub on the second drive automatically that must be done manually

    Not a major issue but does make it hard to boot off drive1 if drive0 failed.

    BIOS would try to boot drive0 or first bootable drive. This would bring up the grub menu which is where it specifies what disk to look at hd0,1. Id have to look at my notes to be sure. So yea it might not be necessary just a force of habit. I specify bootable raid array in raid cards and windows installations so just a force of habit.


    Believe this is reinstalling the grub bootloader on drive0 and drive1 (hd0,0 and hd1,0)
    Code:
    7 Preparing GRUB (Part 1) 
     
    Afterwards we must install the GRUB bootloader on the second hard drive /dev/sdb:
     
    Run these commands:
    grub
     
    On the GRUB shell, type in the following commands:
    grub> root (hd0,0)
     Filesystem type is ext2fs, partition type 0x83
     
    grub> setup (hd0)
     Checking if "/boot/grub/stage1" exists... no
     Checking if "/grub/stage1" exists... yes
     Checking if "/grub/stage2" exists... yes
     Checking if "/grub/e2fs_stage1_5" exists... yes
     Running "embed /grub/e2fs_stage1_5 (hd0)"...  17 sectors are embedded.
    succeeded
     Running "install /grub/stage1 (hd0) (hd0)1+17 p (hd0,0)/grub/stage2 /grub/menu.lst"... succeeded
    Done.
     
    grub> root (hd1,0)
     Filesystem type is ext2fs, partition type 0xfd
     
    grub> setup (hd1)
     Checking if "/boot/grub/stage1" exists... no
     Checking if "/grub/stage1" exists... yes
     Checking if "/grub/stage2" exists... yes
     Checking if "/grub/e2fs_stage1_5" exists... yes
     Running "embed /grub/e2fs_stage1_5 (hd1)"...  17 sectors are embedded.
    succeeded
     Running "install /grub/stage1 (hd1) (hd1)1+17 p (hd1,0)/grub/stage2 /grub/menu.lst"... succeeded
    Done.
     
    grub> quit
    Last edited by Big D; 06-17-2011 at 08:27 AM.
    The beatings shall continue until morale improves!

Page 3 of 3 FirstFirst 123

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  

SEO by vBSEO 3.6.0 PL2