Jump to content
ironicbadger

[XEN VM IMG] ArchVM <--- deprecated 01/07/2014

688 posts in this topic Last Reply

Recommended Posts

I didn't see this anywhere in the thread (if I missed it, my apologies)...

 

sabnzbd from your repository can't use SSL out of the box. python2-pyopenssl is required for this functionality and needs to be installed (pacman -S python2-pyopenssl).

Share this post


Link to post

I didn't see this anywhere in the thread (if I missed it, my apologies)...

 

sabnzbd from your repository can't use SSL out of the box. python2-pyopenssl is required for this functionality and needs to be installed (pacman -S python2-pyopenssl).

 

Assuming you installed this from Ironic's repository when you install via pacman it actually tells you that in order to enable SSL you need to install other packages. Ironic's video walkthrough actually does this as well if I am not mistaken.

 

Share this post


Link to post

Possibly a silly question but if I fire up VMware Player and create an Arch VM, can I then follow the instructions here and download these packages onto it despite not running under Xen/unRAID ?  I'm assuming I can but thought I would check.

 

I'm trying to build myself a sandbox to play around with Arch and these appliances before  moving to 6.o on my only unRAID (prod) server.

Share this post


Link to post

Possibly a silly question but if I fire up VMware Player and create an Arch VM, can I then follow the instructions here and download these packages onto it despite not running under Xen/unRAID ?  I'm assuming I can but thought I would check.

 

I'm trying to build myself a sandbox to play around with Arch and these appliances before  moving to 6.o on my only unRAID (prod) server.

 

Yup, but bear in mind the xen pv drivers are enabled in /etc/mkinitcpio.conf so you may not have network etc until you remove them.

 

nano /etc/mkinitcpio.conf
----------
remove all the xen- stuff from MODULES""

Ctrl + X to exit and save

mkinitcpio -p linux

 

Reboot VM.

Share this post


Link to post

{Edit}

 

I'm testing the ArchVM with openHAB (Home Automation), mySQL (central DB for XBMC) and Air Video. Let me share my findings:

 

1.) mySQL is just working fine - the SQL import was easy - all XBMC content is perfectly reflected.

 

2.) Air Video is just working fine.

 

3.) {NFS mount is only working with a mount -a after a reboot}. Sorry, just saw Ironicbadgers advise in one of his first threads and the same was discussed earlier. Waiting for the next beta4 :-)

 

4.) openHAB seem to be starting at boot:

[root@ArchApplianceVM ~]# systemctl status openhab
openhab.service - openHAB Home Automation Bus
   Loaded: loaded (/usr/lib/systemd/system/openhab.service; enabled)
   Active: active (running) since Tue 2014-02-18 09:16:01 GMT; 49s ago
  Process: 126 ExecStart=/opt/openhab/bin/openhab start (code=exited, status=0/SUCCESS)
Main PID: 292 (wrapper)
   CGroup: /system.slice/openhab.service
           ??292 /opt/openhab/bin/../lib/wrapper /opt/openhab/bin/../etc/wrap...
           ??360 /usr/lib/jvm/java-7-openjdk/jre/bin/java -Dosgi.clean=true -...

Feb 18 09:15:57 ArchApplianceVM openhab[126]: Starting openHAB Home Automati....
Feb 18 09:16:01 ArchApplianceVM openhab[126]: Waiting for openHAB Home Autom....
Feb 18 09:16:01 ArchApplianceVM openhab[126]: running: PID:292
Feb 18 09:16:01 ArchApplianceVM systemd[1]: Started openHAB Home Automation Bus.
Hint: Some lines were ellipsized, use -l to show in full.

....however, only working after a 'systemctl stop openhab', followed by 'systemctl start openhab'

 

...as openHAB is java based I think that issue is around openJDK:

Feb 18 09:17:41 ArchApplianceVM kernel: BUG: Bad page map in process java  pte:0
Feb 18 09:17:41 ArchApplianceVM kernel: addr:00007fe049ce4000 vm_flags:08000070 
Feb 18 09:17:41 ArchApplianceVM kernel: CPU: 0 PID: 744 Comm: java Not tainted 3
Feb 18 09:17:41 ArchApplianceVM kernel:  ffff88007b49e678 ffff88007679dad8 fffff
Feb 18 09:17:41 ArchApplianceVM kernel:  ffff88007679db20 ffffffff8115bf84 00000
Feb 18 09:17:41 ArchApplianceVM kernel:  ffff880076536720 0000000000000320 00007
Feb 18 09:17:41 ArchApplianceVM kernel: Call Trace:
Feb 18 09:17:41 ArchApplianceVM kernel:  [<ffffffff814ed0d3>] dump_stack+0x54/0x
Feb 18 09:17:41 ArchApplianceVM kernel:  [<ffffffff8115bf84>] print_bad_pte+0x1b
Feb 18 09:17:41 ArchApplianceVM kernel:  [<ffffffff8115d349>] vm_normal_page+0x6
Feb 18 09:17:41 ArchApplianceVM kernel:  [<ffffffff8115d84f>] unmap_single_vma+0
Feb 18 09:17:41 ArchApplianceVM kernel:  [<ffffffff8115ed99>] unmap_vmas+0x49/0x
Feb 18 09:17:41 ArchApplianceVM kernel:  [<ffffffff8116812c>] exit_mmap+0x9c/0x1
Feb 18 09:17:41 ArchApplianceVM kernel:  [<ffffffff814f3beb>] ? _raw_spin_lock_i
Feb 18 09:17:41 ArchApplianceVM kernel:  [<ffffffff8105fc19>] mmput+0x59/0x110
Feb 18 09:17:41 ArchApplianceVM kernel:  [<ffffffff81064f4f>] do_exit+0x27f/0xab
Feb 18 09:17:41 ArchApplianceVM kernel:  [<ffffffff810657ff>] do_group_exit+0x3f
Feb 18 09:17:41 ArchApplianceVM kernel:  [<ffffffff810747ad>] get_signal_to_deli
Feb 18 09:17:41 ArchApplianceVM kernel:  [<ffffffff810144d8>] do_signal+0x48/0x9
Feb 18 09:17:41 ArchApplianceVM kernel:  [<ffffffff81014e38>] do_notify_resume+0
Feb 18 09:17:41 ArchApplianceVM kernel:  [<ffffffff814fbeaa>] int_signal+0x12/0x
Feb 18 09:17:41 ArchApplianceVM kernel: Disabling lock debugging due to kernel t
Feb 18 09:17:42 ArchApplianceVM openhab[686]: Stopped openHAB Home Automation Bu
Feb 18 09:17:42 ArchApplianceVM systemd[1]: Stopped openHAB Home Automation Bus.

 

The advise in the openHAB forum was to remove openJDK and to install Oracles JDK7 instead. I saw that in the AUR repository https://aur.archlinux.org/packages/jdk/

 

My question to Ironicbadger is wheather you could transfer this into your repository?

Share this post


Link to post

I'm testing the ArchVM for openHAB (Home Automation)...

Very interested in seeing how you get on with this. I was looking at a different solution, Linux MCE, and didn't know about openHAB. Choices, choices...

Share this post


Link to post

I was curious how this VM solution was working for people. I seem to still be having a few issues with it, and am curious if it's common, or unique to me.

 

1) Plex doesn't seem to find all my shows. I have 135 TV shows, and Plex finds anywhere from 6-100 of them. I've made sure to spin up all my drives and changed directory to the mount point for my UnRAID share, and my last attempt got 100 TV shows. If I do another deep scan I may end up with 70, or 90, and the automatic update and periodic updates don't seem to catch the rest.

2) Sickbeard shows some TV shows as missing their directory, and it also has marked entire shows as ignored (even though it shows the files as downloaded). I had this happen to 15 or so shows (ever episode of every season got marked as ignored).

3) SAB seems really slow for downloading. I have a 150MB internet connection, but SAB runs at 15-20MB even though I've tuned it as per the wiki.

4) SAB/SB will sometimes fail to move the file to the destination folder (presumably because it can't find it)

 

None of the above were issues when I was running these plugins on 5.0, but since moving to 6.0/VM things are not nearly as seamless.

 

I have configured the cron job to change directory to /mnt/TV and mnt/Movies which are both mounts from UnRAID, but I don't know how to confirm they are running (I have the '* * * * * /bin/bash cd /mnt/TV' syntax as per Ironic's first post.)

 

I recognize that we are still running in beta, and that with issues like the 10-30 delay for networking to kick in , and RO NFS shares by default there are going to be some growing pains, but I wanted to get a sense of whether others are having similar issues, or if I am all alone here.

 

Share this post


Link to post

Be sure to follow the naming methodology recommended by Plex. For example:

Sliders/Season 1/Sliders.S01E01.mkv

 

Plex picked up my 21 TV series without issue.

Share this post


Link to post

Be sure to follow the naming methodology recommended by Plex. For example:

Sliders/Season 1/Sliders.S01E01.mkv

 

Plex picked up my 21 TV series without issue.

 

I've maintained this methodology for years. Also, Plex as a 5.0.5 plugin picked up all series no problem. the issue I think is the mount points still don't stick. I had 100 TV shows when I wrote the post, and then it jumped to 110, then a few hours later it was 64, then 107, and now it's 74 (this is all based on Plex polling the shares - I haven't touched anything).

 

Share this post


Link to post

Be sure to follow the naming methodology recommended by Plex. For example:

Sliders/Season 1/Sliders.S01E01.mkv

 

Plex picked up my 21 TV series without issue.

 

I've maintained this methodology for years. Also, Plex as a 5.0.5 plugin picked up all series no problem. the issue I think is the mount points still don't stick. I had 100 TV shows when I wrote the post, and then it jumped to 110, then a few hours later it was 64, then 107, and now it's 74 (this is all based on Plex polling the shares - I haven't touched anything).

I'm having the exact same problem with NFS shares.  I see it not only in Plex, but also Sickbeard where the directory for certain shows shows as "dir is missing".  Even if I try and cd into my NFS folder for a missing show, I get a "directory does not exist" message at the bash prompt.  I went back to SB/SAB/Plex running off of unRAID.  I was having a problem with Crashplan too - I wasn't able to select NFS shares in the webgui to backup.

 

I appreciate all the work involved in providing this VM, and I'm hoping this is all an issue with NFS in b3, and will be fixed in b4.

Share this post


Link to post

Same results here too. I am slowly giving up on autofs in favour of fstab mounts and a script which runs every minute to mount the NFS share when available. Will wait for beta4 tho.

 

Sent from my Nexus 5 using Tapatalk

 

 

Share this post


Link to post

Well, I am glad to hear it's not just me, however this means this is a much larger issue, which sort of sucks too.

 

For the script, is it '* * * * * /bin/bash cd /mnt/UnRAID Mount point'

 

I do have that set, but it doesn't seem to be helping.

 

I will also look into fstab mounts and see if I can figure them out and if it behaves any differently.

 

Unless something dramatically changes this VM solution isn't going to be a viable plug-in replacement.

 

Hopefully we can get beta4 soon and re-test.

Share this post


Link to post

I think it's a little early to give up and declare its not going to work bkastner.

 

The network delay not withstanding the other issues are all fixable, just need to figure out the best way and as I hinted above I working on it. I use this VM every day as my workhorse and so I have a vested interest to get it right.

 

Fwiw fstab mounts are solid in my testing here for 3 days now. Giving a fee more before I declare a winner.

 

Keep the faith!

 

Sent from my Nexus 5 using Tapatalk

 

Share this post


Link to post

I think it's a little early to give up and declare its not going to work bkastner.

 

The network delay not withstanding the other issues are all fixable, just need to figure out the best way and as I hinted above I working on it. I use this VM every day as my workhorse and so I have a vested interest to get it right.

 

Fwiw fstab mounts are solid in my testing here for 3 days now. Giving a fee more before I declare a winner.

 

Keep the faith!

 

Sent from my Nexus 5 using Tapatalk

 

I am still being cautiously optimistic, but it's annoying. :)

 

Personally I have no problem with the network startup delay because it's rare you are going to reboot this once it's running smooth. The loss of connectivity to UnRAID is a major PITA though as I need to manually monitor it now so that I can move files over when SAB fails to. Thankfully with the Olympics last week and this most shows are not broadcasting new, but starting next week this is going to change again, which will make it a lot more work (unless I setup SAB/SB on another pc in parallel for the moment, which I don't really want to do). Also, Plex is pretty useless at the moment, which the wife is unhappy about.

 

Can you share your fstab config? I don't mind trying to set this up, but would rather use yours as a template to make sure the options are right.

 

Share this post


Link to post

I think it's a little early to give up and declare its not going to work bkastner.

 

The network delay not withstanding the other issues are all fixable, just need to figure out the best way and as I hinted above I working on it. I use this VM every day as my workhorse and so I have a vested interest to get it right.

 

Fwiw fstab mounts are solid in my testing here for 3 days now. Giving a fee more before I declare a winner.

 

Keep the faith!

 

Sent from my Nexus 5 using Tapatalk

 

I am still being cautiously optimistic, but it's annoying. :)

 

Personally I have no problem with the network startup delay because it's rare you are going to reboot this once it's running smooth. The loss of connectivity to UnRAID is a major PITA though as I need to manually monitor it now so that I can move files over when SAB fails to. Thankfully with the Olympics last week and this most shows are not broadcasting new, but starting next week this is going to change again, which will make it a lot more work (unless I setup SAB/SB on another pc in parallel for the moment, which I don't really want to do). Also, Plex is pretty useless at the moment, which the wife is unhappy about.

 

Can you share your fstab config? I don't mind trying to set this up, but would rather use yours as a template to make sure the options are right.

 

Just look in the owncloud thread or Google CIFS fstab entry...

 

Sent from my Nexus 5 using Tapatalk

 

 

Share this post


Link to post

 

Just look in the owncloud thread or Google CIFS fstab entry...

 

Sent from my Nexus 5 using Tapatalk

 

Okay, I see the one in the owncloud thread, which I can use as a reference. Since I already have UnRAID in fstab but using NFS is it just moving it to CIFS that is making it better for you?

 

Share this post


Link to post

An update for those who've been experiencing issues with SAB/SB/Plex while using NFS to mount UnRAID in the VM.

 

I switched to CIFS and everything seems to be running smoother. Plex has been consistently finding 133 shows (there are 3 other empty folders it's ignored), and I have not had the bouncing around of episode counts (my tv list would range from 6 to 60 to 110 to 74 to 90, etc).

 

SAB/SB has not processed any new TV shows since I made the change yesterday, but things definitely look more consistent so far.

 

For those interested in switching, here is the updated FSTAB file I have:

 

#

# /etc/fstab: static file system information

#

# <file system> <dir>  <type>  <options>      <dump>  <pass>

#192.168.1.40:/mnt/user/TV          /mnt/TV        nfs rw,rsize=4096,wsize=4096,hard,intr,async,nodev,nosuid 0 0

#192.168.1.40:/mnt/user/Movies    /mnt/Movies  nfs rw,rsize=4096,wsize=4096,hard,intr,async,nodev,nosuid 0 0

//192.168.1.40/TV                        /mnt/TV          cifs rw,guest

//192.168.1.40/Movies                  /mnt/Movies    cifs rw,guest

 

I have the hashed out NFS mounts, and then new CIFS mount points listed.

 

I will have to keep watch over the next few days and make sure SAB/SB are happy, but this definitely seems to be a better config than with NFS.

Share this post


Link to post

So just to clarify the issue is not related to auto fs unmounting after timeout period but also using nfs in fstab didn't work?

Share this post


Link to post

An update for those who've been experiencing issues with SAB/SB/Plex while using NFS to mount UnRAID in the VM.

 

I switched to CIFS and everything seems to be running smoother. Plex has been consistently finding 133 shows (there are 3 other empty folders it's ignored), and I have not had the bouncing around of episode counts (my tv list would range from 6 to 60 to 110 to 74 to 90, etc).

 

SAB/SB has not processed any new TV shows since I made the change yesterday, but things definitely look more consistent so far.

 

For those interested in switching, here is the updated FSTAB file I have:

 

#

# /etc/fstab: static file system information

#

# <file system> <dir>  <type>  <options>      <dump>  <pass>

#192.168.1.40:/mnt/user/TV          /mnt/TV        nfs rw,rsize=4096,wsize=4096,hard,intr,async,nodev,nosuid 0 0

#192.168.1.40:/mnt/user/Movies    /mnt/Movies  nfs rw,rsize=4096,wsize=4096,hard,intr,async,nodev,nosuid 0 0

//192.168.1.40/TV                        /mnt/TV          cifs rw,guest

//192.168.1.40/Movies                  /mnt/Movies    cifs rw,guest

 

I have the hashed out NFS mounts, and then new CIFS mount points listed.

 

I will have to keep watch over the next few days and make sure SAB/SB are happy, but this definitely seems to be a better config than with NFS.

 

Well, I have a "Permission denied" when trying to create a file with anything different than root :

 

touch test
touch: cannot touch ‘test’: Permission denied

 

 

EDIT : Had to file_mode=0777,dir_mode=0777... Not really safe, but well...

Share this post


Link to post

So just to clarify the issue is not related to auto fs unmounting after timeout period but also using nfs in fstab didn't work?

 

This was my experience, yes. NFS mounting period had issues for me, but CIFS seems to be happy so far.

Share this post


Link to post

An update for those who've been experiencing issues with SAB/SB/Plex while using NFS to mount UnRAID in the VM.

 

I switched to CIFS and everything seems to be running smoother. Plex has been consistently finding 133 shows (there are 3 other empty folders it's ignored), and I have not had the bouncing around of episode counts (my tv list would range from 6 to 60 to 110 to 74 to 90, etc).

 

SAB/SB has not processed any new TV shows since I made the change yesterday, but things definitely look more consistent so far.

 

For those interested in switching, here is the updated FSTAB file I have:

 

#

# /etc/fstab: static file system information

#

# <file system> <dir>  <type>  <options>      <dump>  <pass>

#192.168.1.40:/mnt/user/TV          /mnt/TV        nfs rw,rsize=4096,wsize=4096,hard,intr,async,nodev,nosuid 0 0

#192.168.1.40:/mnt/user/Movies    /mnt/Movies  nfs rw,rsize=4096,wsize=4096,hard,intr,async,nodev,nosuid 0 0

//192.168.1.40/TV                        /mnt/TV          cifs rw,guest

//192.168.1.40/Movies                  /mnt/Movies    cifs rw,guest

 

I have the hashed out NFS mounts, and then new CIFS mount points listed.

 

I will have to keep watch over the next few days and make sure SAB/SB are happy, but this definitely seems to be a better config than with NFS.

 

Well, I have a "Permission denied" when trying to create a file with anything different than root :

 

touch test
touch: cannot touch ‘test’: Permission denied

 

 

EDIT : Had to file_mode=0777,dir_mode=0777... Not really safe, but well...

 

I have seen that syntax suggested elsewhere for CIFS as well. I tried with just the basic config I mentioned and touched a file successfully so left it. As for your safety concerns you could try 666 which would mark it r/w for everyone, but not execute.

 

 

Share this post


Link to post

Since I was having such good success with CIFS and my TV mount I tried adding my Movies to Plex. It managed to pick up all 2266 on the first try - even without me refreshing the mount point or spinning up all my drives. I know that I would have only gotten a small fraction with the NFS mount point.

 

CIFS definitely seems to be a more stable solution, and I am back to being confident in my VM solution being viable. I still think it's weird that I need to use CIFS to get 2 Linux boxes to talk correctly, but at least it seems to be working flawlessly.

 

Share this post


Link to post

Looking forward to testing this out after work....

 

I did just realize I still had the cron job configured to cd into each directory (I had that running as part of the NFS testing). I've disabled cron now and will see how things go. It's still trying to find artwork for the movies, but once done I will let it sit for a while and redo the deep scan and see what happens.

 

Even if I have to enable the cron job again I am happy to have a properly working VM at this point. I know this may all change with later betas but I am happy to appear to have proper functionality again without having to resort back to plugins.

Share this post


Link to post

Since I was having such good success with CIFS and my TV mount I tried adding my Movies to Plex. It managed to pick up all 2266 on the first try - even without me refreshing the mount point or spinning up all my drives. I know that I would have only gotten a small fraction with the NFS mount point.

 

CIFS definitely seems to be a more stable solution, and I am back to being confident in my VM solution being viable. I still think it's weird that I need to use CIFS to get 2 Linux boxes to talk correctly, but at least it seems to be working flawlessly.

 

I could have sworn I said a few pages ago that we should consider NFS broken in beta3 and that there was nothing I could do about it! Hint: I definitely said that  ;)

 

This is some good 'testing' you're all undertaking with NFS vs Samba (CIFS). I'm sure Tom will be looking at these issues even if not posting, who knows. Keep plucking away chaps!

 

Anyway, I'm glad CIFS is working out for y'all! :)

Share this post


Link to post
Guest
This topic is now closed to further replies.