• unRAID OS version 6.5.1-rc3 available


    limetech
    • Closed

    Welcome to the new experimental Prerelease Support board!  We have added the ability to tag issues/bugs with a Priority and mark when solved.  Please report only issues/bugs which are new in the prerelease.

     

    Too many issues with docker 17.12.x release, and we haven't had enough internal testing with 18.03.x release, so we are reverting back to 17.09.1.  We are anticipating moving to 18.03.x in unRAID OS 6.6.

     

    Also: got rid of the "wont fix" tag and replaced with "Retest". If you see this on your bug report it means that you should see if the latest release fixes the problem.  This might seem odd that we, the developers, might not know this; but, often we cannot reproduce many issues that are reported.  Some issues end up being solved by a newer kernel release.  Some issues we may have developed a theory why it might be happening and then put a change in the code to address it.

     


    Version 6.5.1-rc3 2018-03-30

    Summary:

    • Security updates, bug fixes and UI improvements.

    Base distro:

    • docker: version 17.09.1 (downgrade from version 17.12.1)
    • openssl: version 1.0.2o (CVE-2018-0739)
    • openssl-solibs: version 1.0.2o (CVE-2018-0739)

    Linux kernel:

    • version 4.14.31
      • with tcp_reset_sk_send_head_in_tcp_write_queue_purge patch
      • with tehuti pci-id [1fc9:3015] patch (experimental)

    Management:

    • Ensure hostname contains only valid NETBIOS characters
    • Yet another mover logging fix
    • webgui: Fixed display of error codes in SWAL
    • webgui: VMmachine updates
      • changed DETACH is available when VM is running
      • default bus to VirtIO
      • minor layout changes
    • webgui: Docker only remove icons when container/image is deleted
    • webgui: DockerClient: sort container names alphabetically and port numbers numerically
    • webgui: Sort unassigned devices by disk size, disk model, disk name
      • This groups disks together by size, followed by same model, and lastly their assigned name (sdX)
    • webgui: Dokcer combine used ports and ips in single list
    • webgui: Fixed: dashboard apps, properly initialize menu variable
    • webgui: Reduced container list to one call to docker only for all containers at once
    • webgui: Fixed VM 'detach' button
    • webgui: Add short delay in array list update after spin up/spin down of disk
    • webgui: Fix Create Docker php warning
    • webgui: Diagnostic improvements
    • webgui: Fixed false "next" announcement when on "stable " release
    • webgui: Fixed removal of container and updating of json file
    • webgui: Restore Apply button to previous behavior

    Edited by limetech

    • Like 1



    User Feedback

    Recommended Comments

    Curious how this is being handled

    Quote
    • Ensure hostname contains only valid NETBIOS characters

    I purposely named my server to be incompliant (Server_A), and I see in the syslog that it's been automatically redone as Server-A (although the top right of the UI) still states Server_A.  

     

    Server_A's webUI cannot be reached, but Server-A can be.  Don't want the nitty gritty, but an overview of the rules which you're applying.

    Link to comment

    Again, this should only affect a very small set of users, only those who somehow set an invalid netbios server name (and webgui will not permit this).  What the code does is not change the string set in the ident.cfg file, instead it substitutes all invalid chars with a dash character before setting the host name:

     

    # limetech - read our ident.cfg file
    NAME="Tower"
    timeZone="America/Los_Angeles"
    if [ -r /boot/config/ident.cfg ]; then
      /usr/bin/fromdos </boot/config/ident.cfg >/var/tmp/ident.cfg
      source /var/tmp/ident.cfg
      NAME=${NAME//[^a-zA-Z\-\.0-9]/\-}
    fi
    
    # Set the hostname.
    # limetech - get hostname from ident.cfg file on flash and ensure hostname is
    # defined as localhost alias in /etc/hosts (this lets wins name resolution work)
    /bin/hostname $NAME
    /bin/hostname >/etc/hostname
    echo "# Generated" >/etc/hosts
    echo "127.0.0.1 $NAME localhost" >>/etc/hosts
    

     

    Link to comment
    1 minute ago, limetech said:

    instead it substitutes all invalid chars with a dash character

    Exactly what I wanted to know.  Helps out the helpers around here knowing the rules being applied since its not obvious

     

    Link to comment
    3 minutes ago, Squid said:

    Exactly what I wanted to know.  Helps out the helpers around here knowing the rules being applied since its not obvious

     

    We didn't want to write the "corrected' server name back to the ident.cfg file because then if user downgraded their server name would still be wrong, ie, not what they set it to.

    Link to comment
    Quote

    webgui: Sort unassigned devices by disk size, disk model, disk name

    • This groups disks together by size, followed by same model, and lastly their assigned name (sdX)

     

    Love it! Works fantastic! Thanks @bonienl ! I did a 'new config' just to try it out xD

     

    Link to comment

    @limetech

    You may want to consider this

    webgui: VMmachine updates
    
        default bus to VirtIO

    for FreeBSD because it doesn't work, at least in opnsense and pfsense. With Sophos XG doesn't work either although it is based on linux. In case this is the bus for the hardrive image, I had to use SATA.

    Link to comment

    This change sets the bus type to VirtIO when it is empty. Normally a user would choose a bus type, like in your case SATA.

     

    Link to comment

    I notice that in -rc3 the SMART reports included in the diagnostics zip no longer end with .txt which makes them less likely to be opened by text editors with one click. Is it intentional?

    Link to comment

    I updated and now I can't get a listing for My Apps or Plugins. 

    I just got the following message. 

     

    ************************************

     

    Download of appfeed failed.

    Community Applications requires your server to have internet access. The most common cause of this failure is a failure to resolve DNS addresses. You can try and reset your modem and router to fix this issue, or set static DNS addresses (Settings - Network Settings) of 8.8.8.8 and 8.8.4.4 and try again.

    Alternatively, there is also a chance that the server handling the application feed is temporarily down. Switching CA to operate in Legacy Mode might temporarily allow you to still utilize CA.

    Last JSON error Recorded: JSON Error: Syntax error

     

    ***************************************

     

    After this message I did a test ping to another site and it came back as well I did an update to all my Dockers just fine. I just get get a listing for Apps or Plugins since this update. 

    Link to comment
    3 hours ago, kizer said:

    I updated and now I can't get a listing for My Apps or Plugins. 

     

    See here:

     

    Link to comment
    10 hours ago, John_M said:

    I notice that in -rc3 the SMART reports included in the diagnostics zip no longer end with .txt which makes them less likely to be opened by text editors with one click. Is it intentional?

     

    No, a regression error. Fixed in next release.

     

    Thanks for reporting.

    Edited by bonienl
    • Like 1
    Link to comment
    53 minutes ago, bonienl said:

     

    No, a regression error. Fixed in next release.

     

    Thanks for reporting.

    :(  That one's on me :sniff: :sniff:

    Link to comment
    4 hours ago, John_M said:

     

    See here:

     

    @John_M Thanks for the Info. I was totally not wanting to spend the AM trying to fix this. 

    Link to comment

    I'm really happy that some issues got reported and brought to developers attention.

    HOWEVER

    Please stop using the Announcement topic to report issues.  We're doing the best we can to make it easy to report issues right here in this forum.  It is simply too hard to track this stuff if otherwise.

    My next step is to lock these announce topics and force people to create issue reports.

    The next step after that is to forget about the forum and use Bugzilla or Mantis.  We have resisted that because if forces people to create multiple accounts which means more effort.  We have found the best way to find out about issues is to make it as dirt simple as possible to report them.  But I have to draw the line here: don't report bugs in the announcement posts.

    • Like 3
    Link to comment

    Perhaps have a second post in the locked thread that asks people to "Like" the second post it if they're not experiencing any issues.  That way it's possible to track the number of forum posters who are not experiencing issues garnering feed back without cluttering the forums.

     

    I'm sure there are many people that wait until a couple people report in that the latest testing release is stable before they install for themselves.

    • Like 4
    Link to comment


    Guest
    This is now closed for further comments

  • Status Definitions

     

    Open = Under consideration.

     

    Solved = The issue has been resolved.

     

    Solved version = The issue has been resolved in the indicated release version.

     

    Closed = Feedback or opinion better posted on our forum for discussion. Also for reports we cannot reproduce or need more information. In this case just add a comment and we will review it again.

     

    Retest = Please retest in latest release.


    Priority Definitions

     

    Minor = Something not working correctly.

     

    Urgent = Server crash, data loss, or other showstopper.

     

    Annoyance = Doesn't affect functionality but should be fixed.

     

    Other = Announcement or other non-issue.