unRAID Server Release 6.1-rc6 Available

[limetech]

If you are running 6.1-rc3/-rc4/-rc5 you should be able to click Check for Updates on the Plugins page to get this release.

 

Alternately, navigate to Plugins/Install Plugin and paste the following link into the box and click Install:

 

https://raw.githubusercontent.com/limetech/unRAIDServer-6.1-rc/master/unRAIDServer.plg

 

Plugin authors: if you haven't already, please read this topic regarding command execution:

http://lime-technology.com/forum/index.php?topic=42259.msg401951#msg401951

 

New users to 6.1-rc: please disable/enable notifications to get them working again.

 

All  users: you should hit F5 in browser window or clear browser cache after updating.

 

---

 

unRAID Server Version 6.1-rc6 Release Notes
===============================================

Security Advisory updates (6.1-rc1)
-----------------------------------
- ntp: 4.2.8p3 (SSA:2015-188-03)
- openssl: 1.0.1p (SSA:2015-190-01)

Security Advisory updates (6.1-rc3)
-----------------------------------
- php: 5.4.43 (SSA:2015-198-02)

Security Advisory updates (6.1-rc6)
-----------------------------------
- gnutls: 3.3.17.1 (SSA:2015-233-01)

Other changes (6.1-rc1)
-----------------------
- btrfs-progs: version 4.1.0
- docker: version 1.7.1
- docker: less verbose logging by default
- emhttp: fix: let Pro start regardless of attached device count
- emhttp: bridge forward delay configurable
- emhttp: support disk share export Yes/No/Auto
- libvirt: version 1.2.17
- linux: version 4.1.1
- linux: support "Lite-On IT Corp. / Plextor M6e PCI Express SSD [Marvell 88SS9183] (rev 14)", see https://lkml.org/lkml/2015/2/2/226
- ntp: enable by default
- php: version 5.4.42
- php: enable pcntl
- smb: include linux "hidden" objects in directory listings
- unRAIDServer.plg: improvements in unRAID-5 update: ping google instead of local gateway to determine if network up
- vm.manager: support OpenELEC template
- webGui: include linux "hidden" objects in directory listings
- webGui: integrate latest dynamix changes; many changes, refer to github
- webGui: install key: accept pre-keyserver links to key files
- webGui: rearrange columns on Main page
    
Other changes (6.1-rc2)
-----------------------
- emhttp: properly handle non-rotational devices standby mode (ie, 'spindown')
- emhttp: require credentials on all accesses, not just webGui pages
- emhttp: get rid of passing arbitrary commands in URL
- webGui: added "Hide dot files" control on SMB Settings page
- webGui: fix issues with uploading thumbnails, banner
- webGui: fix notifications not initialized on reboot
- webGui: fix docker update regression introduced in 6.1-rc1

Other changes (6.1-rc3)
-----------------------
- avahi: supress "Invalid response packet from host" messages
- btrfs-progs: version 4.1.2
- emhttp: fix spindown regression introduced in -rc2.
- emhttp: fix problem not able to start multiple btrfs scrubs
- libvirt: version 1.2.18
- linux: version 4.1.5
- php: enable iconv per user request
- samba: version 4.1.19
- webGui: all '#command' and 'cmd' values must reference scripts within /usr/local/emhttp/ jail.
- webGui: fix image upload issues
- webGui: Disable Array Stop button when mover is running
- webGui: Fixed memory max display in dashboard
- webGui: VM Manger: Updated noVNC (using latest as of 8/5/15, commit 8f3c0f6)
- webGui: Replace alert-box for SweetAlert (enhanced messaging)
- webGui: additional options for text/graphical display of Used/Free.
- webGui: Add APCUPS summary to dashboard page
- webGui: other misc. bug fixes
- OpenELEC VM template: Fix: allow spaces and quotes in download folder
- OpenELEC: added 6.0-beta3 image (beta3 includes hdmi sound fixes for nvidia cards and corrected 'addon_url' to allow plugins in Kodi once again)
- OpenELEC: use readonly for openelec disk image

Other changes (6.1-rc4)
-----------------------
- webGui: fix monitor/notification regressions resulting from closing 'arbitrary command execution' holes

Other changes (6.1-rc5)
-----------------------
- webGui: fix another regression resulting from closing 'arbitrary command execution' holes
- webGui: integrated eric's feedback for kool kids

Other changes (6.1-rc6)
-----------------------
- apcupsd: fix regression: 'powerdown' script not invoked correctly
- emhttp: properly handle requests from localhost (IP address 127.0.0.1)
- emhttp: let Trial support disks in any slots
- mover: only move objects where "Use Cache" is set to "Yes" for a share
- php: version 5.4.44
- php: remove iconv (not needed after all)
- slack: ca-certificates version 20150426
- slack: check USB Flash device upon startup (set 6th field of fstab to 1)
- webGui: use 'ftpusers' script to define ftp users
- webGui: use 'agent' script to control notification agents
- webGui: accept only properly formed 'cmd=' and '#command' query variables
- webGui: other misc. bug fixes and improvements

[Squid]

Plugin authors: if you haven't already, please read this topic regarding command execution:

http://lime-technology.com/forum/index.php?topic=42259.msg401951#msg401951

ok.  I'm a little bit confused here now.

 

Under RC-5 I was receiving zero warning about deprecations within the log for my openBox() paths. Beyond the system logging that I was actually running the routine.

 

Aug 24 17:37:34 Server_A emhttp: /usr/local/emhttp/plugins/community.applications/scripts/showInfo.php 27 2>&1
Aug 24 17:37:45 Server_A emhttp: /usr/local/emhttp/plugins/community.applications/scripts/showDesc.php 53 2>&1
Aug 24 17:37:48 Server_A emhttp: /usr/local/emhttp/plugins/community.applications/scripts/showInfo.php 67 2>&1

 

Under RC-6 running the same commands, openBox() won't work properly and I'm getting this logged:

 

Aug 24 17:44:13 Server_A emhttp: run_cmd: invalid cmd: /plugins/community.applications/scripts/showDesc.php 125

 

Viewing the source of my page shows this:

 

  openBox('/plugins/community.applications/scripts/showDesc.php ' + appNumber,appName + " Description",400,400);

 

I thought that this was the correct format to use (it certainly ran without any warnings under RC5).  Haven't checked all of the operations of CA yet....

 

EDIT:  The scripts that CA calls for dockerfile conversions are still operational...

 

Using plugins/$plugin/scripts/... Doesn't work.

Using /usr/local/emhttp/plugins/$plugin/scripts/... doesn't work

Using emhttp/plugins/$plugin/scripts doesn't work

Using /emhttp/plugins/$plugin/scripts doesn't work

 

Actually, as far as I can tell, no path to openBox works...  :'(

 

I was already looking at switching over to sweetAlert for the openBox routines, but that carries some baggage as sweetAlert doesn't place a scroll bar if the text doesn't fit

 

[pickthenimp]

 

- mover: only move objects where "Use Cache" is set to "Yes" for a share

[/code]

 

Out of curiosity...could this resolve the issue I have been having where my server locks up every night when it runs the mover script?

 

https://lime-technology.com/forum/index.php?topic=42319.msg403458#msg403458

[RobJ]

Users also need to re-apply the scheduled parity check settings (go to Settings -> Scheduler and re-apply in the Parity Check section).  If not done, the scheduled parity check will not happen.

 

Reminder: if disk shares suddenly stop working, please reread the v6.1-rc1 announcement post.

[trurl]

 

- mover: only move objects where "Use Cache" is set to "Yes" for a share

[/code]

 

Out of curiosity...could this resolve the issue I have been having where my server locks up every night when it runs the mover script?

 

https://lime-technology.com/forum/index.php?topic=42319.msg403458#msg403458

If you already have cache-only for those shares you don't want moved then it won't have any effect. The point of this change to mover is to allow creation of cache shares that won't get moved by default. With this change, a share created on cache won't be moved unless you explicitly set it to cache-yes.

[limetech]

Plugin authors: if you haven't already, please read this topic regarding command execution:

http://lime-technology.com/forum/index.php?topic=42259.msg401951#msg401951

ok.  I'm a little bit confused here now.

 

Under RC-5 I was receiving zero warning about deprecations within the log for my openBox() paths. Beyond the system logging that I was actually running the routine.

 

Aug 24 17:37:34 Server_A emhttp: /usr/local/emhttp/plugins/community.applications/scripts/showInfo.php 27 2>&1
Aug 24 17:37:45 Server_A emhttp: /usr/local/emhttp/plugins/community.applications/scripts/showDesc.php 53 2>&1
Aug 24 17:37:48 Server_A emhttp: /usr/local/emhttp/plugins/community.applications/scripts/showInfo.php 67 2>&1

 

Under RC-6 running the same commands, openBox() won't work properly and I'm getting this logged:

 

Aug 24 17:44:13 Server_A emhttp: run_cmd: invalid cmd: /plugins/community.applications/scripts/showDesc.php 125

 

Viewing the source of my page shows this:

 

  openBox('/plugins/community.applications/scripts/showDesc.php ' + appNumber,appName + " Description",400,400);

 

I thought that this was the correct format to use (it certainly ran without any warnings under RC5).  Haven't checked all of the operations of CA yet....

 

EDIT:  The scripts that CA calls for dockerfile conversions are still operational...

 

Using plugins/$plugin/scripts/... Doesn't work.

Using /usr/local/emhttp/plugins/$plugin/scripts/... doesn't work

 

I was already looking at switching over to sweetAlert for the openBox routines, but that carries some baggage as sweetAlert doesn't place a scroll bar if the text doesn't fit

 

The problem is you can't pass arguments directly with commands like we used to allow.

This line:

  openBox('/plugins/community.applications/scripts/showDesc.php ' + appNumber,appName + " Description",400,400);

has to be something like this:

  openBox('/plugins/community.applications/scripts/showDesc.php?app=' + appNumber,appName + " Description",400,400);

 

We have to enforce this or else it would be possible to execute compound commands and back-tick'ed commands since ultimately php's exec() function is called (or one of it's brothers such as shellexec()) and these functions always invoke a shell, in our case bash.

 

Edit: this wasn't marked deprecated in -rc5  because we didn't realize that vulnerability then.  It was during -rc6 development I realized this problem.  (I had read somewhere that php's exec() did not call the shell, that it operated like C execv() fucntion - but that is not the case it turns out.)

[dlandon]

Clicking on "Reset" in the UPS Settings page sets the "Start APC UPS Daemon" to 'No' but does not stop the UPS daemon if it is already running.  On reboot, the "Start APC UPS Daemon" is set to 'No', but the daemon starts anyway.

[limetech]

Clicking on "Reset" in the UPS Settings page sets the "Start APC UPS Daemon" to 'No' but does not stop the UPS daemon if it is already running.  On reboot, the "Start APC UPS Daemon" is set to 'No', but the daemon starts anyway.

 

You mean clicking on 'Default' does this..?  That's what I see.

[dlandon]

Clicking on "Reset" in the UPS Settings page sets the "Start APC UPS Daemon" to 'No' but does not stop the UPS daemon if it is already running.  On reboot, the "Start APC UPS Daemon" is set to 'No', but the daemon starts anyway.

 

You mean clicking on 'Default' does this..?  That's what I see.

 

Sorry.  Yes 'Default'.

[jbartlett]

One potential thing to add to the release notes: If you receive access denied errors on shares that have access rights enabled but you have access, reboot the Windows PC you received the access denied error on.

[RobJ]

A discovery, I noticed on reboot after update to 6.1-rc6 that on the console were messages about fsck.fat, running on the flash drive.  I've looked and it's not logged, in syslog or dmesg, so only way you will see it is on an attached console (or perhaps IPMI console?).  It would be nice if there's a way to log its output.

 

I do see "slack: check USB Flash device upon startup (set 6th field of fstab to 1)" among the rc6 improvements.

 

On mine, it found 2 problems, "0x41: Dirty bit is set.  Fs was not properly unmounted ..." and "There are differences between boot sector and its backup".  Both were automatically fixed, so no more "kernel: FAT-fs (sda1): Volume was not properly unmounted. Some data may be corrupt. Please run fsck".

 

The first issue we have known about, as it still happens, even though there's a sync command after powerdown saves the syslog.  There must still be a last write to flash that's not flushed before reboot.  The second is something I haven't seen before, makes me wonder how well syslinux did its job.  Hopefully doesn't repeat on next reboot.

[limetech]

A discovery, I noticed on reboot after update to 6.1-rc6 that on the console were messages about fsck.fat, running on the flash drive.  I've looked and it's not logged, in syslog or dmesg, so only way you will see it is on an attached console (or perhaps IPMI console?).  It would be nice if there's a way to log its output.

 

I do see "slack: check USB Flash device upon startup (set 6th field of fstab to 1)" among the rc6 improvements.

 

On mine, it found 2 problems, "0x41: Dirty bit is set.  Fs was not properly unmounted ..." and "There are differences between boot sector and its backup".  Both were automatically fixed, so no more "kernel: FAT-fs (sda1): Volume was not properly unmounted. Some data may be corrupt. Please run fsck".

 

The first issue we have known about, as it still happens, even though there's a sync command after powerdown saves the syslog.  There must still be a last write to flash that's not flushed before reboot.  The second is something I haven't seen before, makes me wonder how well syslinux did its job.  Hopefully doesn't repeat on next reboot.

 

RobJ always impressed with your attention to detail!!

 

Yes we set the 6th field in fstab entry for /boot mounted on the usb flash to a "1". Previously it was 0.  When set, this tells 'mount' to execute a fsck on the flash file system before mounting it.  In this case it calls 'fsck.vfat'.  The fsck is called with -a option which says to automatically fix errors/warnings it might find.

 

After quite a bit of investigation it seems that syslinux does not write the 'backup' boot sector after writing the MBR as part of creating a bootable usb flash device.  Most people will therefore see this boot sector backup being "fixed" once.  After that it should not happen again.

 

As for the 'dirty bit being set', that means system reset occurred before flash was un-mounted.  I don't see this happen on our test servers so I'd be interested in knowing how you can make this happen in "normal operation".

[CHBMB]

Well, you'll be disappointed by my attention to detail but all seems to be well here.

 

I think I understood less than five percent of the last two posts...

[bonienl]

Not explicitly mentioned, but the dashboard page has now additional indicators to show in real-time the utilization of the flash drive, /var/log location and docker image.

 

Also a first warning notification will be sent when the docker image utilization reaches the disk warning level and a subsequent alert notification when the disk critical level is reached (both these settings can be adjusted under Display Settings).

 

[bonienl]

Clicking on "Reset" in the UPS Settings page sets the "Start APC UPS Daemon" to 'No' but does not stop the UPS daemon if it is already running.  On reboot, the "Start APC UPS Daemon" is set to 'No', but the daemon starts anyway.

 

You mean clicking on 'Default' does this..?  That's what I see.

 

Sorry.  Yes 'Default'.

 

The default settings are not properly copied over. I have made a correction.

 

[bkastner]

I just upgraded to rc6 and received an email notification that my Docker image disk utilization is at 81%. Very handy notification, but it raises the question on how I should go about addressing this.

 

I apologize if this is not rc6 related per se, but since it seems to be a rc6 addition I am guessing others may come here once seeing the notification as well.

[bonienl]

I just upgraded to rc6 and received an email notification that my Docker image disk utilization is at 81%. Very handy notification, but it raises the question on how I should go about addressing this.

 

I apologize if this is not rc6 related per se, but since it seems to be a rc6 addition I am guessing others may come here once seeing the notification as well.

 

It is very easy to increase the size of the Docker image file.

 

Under settings -> docker first disable docker, this will allow you to enter a new size value. E.g. change 10G to 20G and re-enable docker.

 

[bkastner]

I just upgraded to rc6 and received an email notification that my Docker image disk utilization is at 81%. Very handy notification, but it raises the question on how I should go about addressing this.

 

I apologize if this is not rc6 related per se, but since it seems to be a rc6 addition I am guessing others may come here once seeing the notification as well.

 

It is very easy to increase the size of the Docker image file.

 

Under settings -> docker first disable docker, this will allow you to enter a new size value. E.g. change 10G to 20G and re-enable docker.

 

It will just expand? I always thought that would create a new container and require me to redownload the Dockers.

[bonienl]

I just upgraded to rc6 and received an email notification that my Docker image disk utilization is at 81%. Very handy notification, but it raises the question on how I should go about addressing this.

 

I apologize if this is not rc6 related per se, but since it seems to be a rc6 addition I am guessing others may come here once seeing the notification as well.

 

It is very easy to increase the size of the Docker image file.

 

Under settings -> docker first disable docker, this will allow you to enter a new size value. E.g. change 10G to 20G and re-enable docker.

 

It will just expand? I always thought that would create a new container and require me to redownload the Dockers.

 

It will just expand, without creating anything new or loosing anything existing.

 

[bkastner]

I just upgraded to rc6 and received an email notification that my Docker image disk utilization is at 81%. Very handy notification, but it raises the question on how I should go about addressing this.

 

I apologize if this is not rc6 related per se, but since it seems to be a rc6 addition I am guessing others may come here once seeing the notification as well.

 

It is very easy to increase the size of the Docker image file.

 

Under settings -> docker first disable docker, this will allow you to enter a new size value. E.g. change 10G to 20G and re-enable docker.

 

It will just expand? I always thought that would create a new container and require me to redownload the Dockers.

 

It will just expand, without creating anything new or loosing anything existing.

 

Great to know. I just tried it and it worked great. I also got an updated notification that my disk utilization has returned to normal.

 

Great addition! I much prefer finding out that I am running out of space this way, verses when there's an actual issue.

[Guzzi]

Not explicitly mentioned, but the dashboard page has now additional indicators to show in real-time the utilization of the flash drive, /var/log location and docker image.

Hi bonienl,

nice addition - but there might be a little bug there in case, no dockers are used (and no docker image exists) - in that case it always shows "20%" - I'd suggest to simply not display docker image usage in case there is none!?

[dlandon]

The Docker Settings page does not refresh after doing a "Scrub".  I have to manually refresh the page with the browser refresh button.

[eschultz]

Not explicitly mentioned, but the dashboard page has now additional indicators to show in real-time the utilization of the flash drive, /var/log location and docker image.

Hi bonienl,

nice addition - but there might be a little bug there in case, no dockers are used (and no docker image exists) - in that case it always shows "20%" - I'd suggest to simply not display docker image usage in case there is none!?

That's not a bug, it's a feature!  It doubles as ram % usage when docker is disabled

[bonienl]

The Docker Settings page does not refresh after doing a "Scrub".  I have to manually refresh the page with the browser refresh button.

 

Yeah manual refreshing is required at this moment. To make that clearer a "refresh" button is added in the next release.

 

[Guzzi]

Not explicitly mentioned, but the dashboard page has now additional indicators to show in real-time the utilization of the flash drive, /var/log location and docker image.

Hi bonienl,

nice addition - but there might be a little bug there in case, no dockers are used (and no docker image exists) - in that case it always shows "20%" - I'd suggest to simply not display docker image usage in case there is none!?

That's not a bug, it's a feature!  It doubles as ram % usage when docker is disabled

Are you sure? As RAM usage is already in the row right above and show different value. And even IF that is the case, then the text would have to be adapted, as it still shows "flash : log : docker" even if there is no docker enabled and no docker image file present.