unRAID Server release 4.5-beta11 available

[limetech]

 

That fixed the symlink problem with Snow Leopard

 

I'll check on my media player later, I'm late for work right now,

 

Running on 4.5beta11 now...  My TVIX media player still can't see symlinks over NFS, but can see them over Samba?

 

Mark.

 

Try a 'relative symlink'.  For example, suppose you have this:

 

/mnt/disk1/Movies

/mnt/disk1/Photos

/mnt/disk1/data

 

Which implies these user shares:

 

/mnt/user/Movies

/mnt/user/Photos

/mnt/user/data

 

And via linux console you type this:

 

cd /mnt/disk1/data

ln -s /mnt/user/Movies .

ln -s /mnt/user/Photos .

 

This would create symlinks so that via /mnt/user/data you would see:

 

/mnt/user/data/Movies

/mnt/user/data/Photos

 

I believe this is what you are doing now, correct??

 

If so, try creating the symlinks like this:

 

cd /mnt/disk1/data

ln -s ../../user/Movies .

ln -s ../../user/Photos .

 

This makes the symlinks "relative".  Works via Windows & Mac - let me know if this works with your TVIX.

 

 

[limetech]

BETA 10 Torrent deleted from linuxtracker... here's beta 11 torrent:

 

http://linuxtracker.org/index.php?page=torrent-details&id=372dfcd849e19024605dcd434917948471f89e97

 

Cheers,

Matt

 

Thanks (to you and Weebo) for setting up the Torrents, much appreciated!

[limetech]

Does anyone know if the paused I/O on drive spin-up issue (where, if you were playing a video from a spinning disk, that video stream would halt if a file on a spun-down disk was accessed, and resume after the disk completely spun up) is fixed in this release?

 

 

Do you have a syslog from when this would happen?

[limetech]

4.5-beta11 still not any faster for me with regards to uploading. Only getting 14 MB/s to the cache. unRAID 4.4.2 gives me 38+ MB/s.  

 

That should not be the case - please try -beta11 again, do a transfer, and then post your system log.

[Suse User]

 

That fixed the symlink problem with Snow Leopard

 

I'll check on my media player later, I'm late for work right now,

 

Running on 4.5beta11 now...  My TVIX media player still can't see symlinks over NFS, but can see them over Samba?

 

Mark.

 

Try a 'relative symlink'.  For example, suppose you have this:

 

/mnt/disk1/Movies

/mnt/disk1/Photos

/mnt/disk1/data

 

Which implies these user shares:

 

/mnt/user/Movies

/mnt/user/Photos

/mnt/user/data

 

And via linux console you type this:

 

cd /mnt/disk1/data

ln -s /mnt/user/Movies .

ln -s /mnt/user/Photos .

 

This would create symlinks so that via /mnt/user/data you would see:

 

/mnt/user/data/Movies

/mnt/user/data/Photos

 

I believe this is what you are doing now, correct??

 

If so, try creating the symlinks like this:

 

cd /mnt/disk1/data

ln -s ../../user/Movies .

ln -s ../../user/Photos .

 

This makes the symlinks "relative".  Works via Windows & Mac - let me know if this works with your TVIX.

 

Recreated symlinks using relative technique and still the same problem, symlinks are fine in Windows & Snow Leopard, just not visible on the media player via NFS.

 

User shares & disk shares are visible when I NFS share. With /mnt/disk1 or /mnt/user I see directories, but the 'data' directory appears empty via NFS and populated with accessible symlinks via Samba.

 

Mark.

[brainbone]

Does anyone know if the paused I/O on drive spin-up issue (where, if you were playing a video from a spinning disk, that video stream would halt if a file on a spun-down disk was accessed, and resume after the disk completely spun up) is fixed in this release?

 

 

Do you have a syslog from when this would happen?

 

No, but it was apparently a long standing issue. (see: http://lime-technology.com/forum/index.php?topic=3237.0 )

 

That said, 4.5-beta11 (and possibly 4.5-beta8, though I didn't try it) seems to have resolved this issue that has haunted me since I started using unRAID.  I haven't done extensive testing since switching to 4.5-beta11, but it looks like I may not need to keep my disks spinning 24/7 any longer! Woohoo!

 

[Romir]

Does anyone know if the paused I/O on drive spin-up issue (where, if you were playing a video from a spinning disk, that video stream would halt if a file on a spun-down disk was accessed, and resume after the disk completely spun up) is fixed in this release?

 

 

Do you have a syslog from when this would happen?

Here's a beta11 e1000 syslog where streaming video was interrupted at the end. Spinning up a disk by accessing it while streaming typically, but not always, causes this usb reset entry for me.

 

tower kernel: usb 2-2: reset high speed USB device using ehci_hcd and address 3

[smino]

Wow, thanks for working out all the bugs guys! I was going to upgrade my prod system, but I think I will setup a dev box !

[limetech]

Does anyone know if the paused I/O on drive spin-up issue (where, if you were playing a video from a spinning disk, that video stream would halt if a file on a spun-down disk was accessed, and resume after the disk completely spun up) is fixed in this release?

 

 

Do you have a syslog from when this would happen?

 

No, but it was apparently a long standing issue. (see: http://lime-technology.com/forum/index.php?topic=3237.0 )

 

That said, 4.5-beta11 (and possibly 4.5-beta8, though I didn't try it) seems to have resolved this issue that has haunted me since I started using unRAID.  I haven't done extensive testing since switching to 4.5-beta11, but it looks like I may not need to keep my disks spinning 24/7 any longer! Woohoo!

 

 

Ok, I remember this now & determining that this is a Windows problem.  To make it happen start out with all drives spun-down.  Now start a movie streaming off one of the drives.  There will be a pause of course until the drive spins up.

 

Now while movie is playing, start accessing directories on other disks via same computer & you will see movie pause while other disk spins up.  This is Windows starving it's own media read stream waiting for another directory access stream to finish.  If you go and access directories via another computer, that computer will pause waiting for the directory listings (waiting for disk to spin up) but the movie on the first computer still plays fine.

 

I think the reason this appears as just a momentary glitch when playing music is because the music player buffers enough data that it just starts to empty as the spin up nears completion.

[Stokkes]

Ok, I remember this now & determining that this is a Windows problem.  To make it happen start out with all drives spun-down.  Now start a movie streaming off one of the drives.  There will be a pause of course until the drive spins up.

 

Now while movie is playing, start accessing directories on other disks via same computer & you will see movie pause while other disk spins up.  This is Windows starving it's own media read stream waiting for another directory access stream to finish.  If you go and access directories via another computer, that computer will pause waiting for the directory listings (waiting for disk to spin up) but the movie on the first computer still plays fine.

 

I think the reason this appears as just a momentary glitch when playing music is because the music player buffers enough data that it just starts to empty as the spin up nears completion.

 

Tom,

 

I see this issue on my Mac Mini connected to unRAID via SMB (Using XBMC/Plex). There is the obvious pause when you start playing a media file from a disk that is spun down (which is understandable). While playing a media file, if I'm browsing directories on unRAID or access another disk that has to be spun up, my media will pause for a few seconds (sometimes it's not even a second, sometimes it's 1-2 seconds).

 

Just letting you know.

[lewcass]

Beta 10 apparently makes my Marvell gigabit issue worse.

 

Appears to be some factor other than the NIC drivers at play as they supposedly have not changed.

 

With beta 10 I am unable to obtain a gigabit connection with my primary on-board NIC regardless of whether the secondary NIC is enabled.

With beta 8 I get gigabit only if the secondary is enabled.

With beta 6 I get gigabit only if the secondary is disabled.

 

Motherboard is Asus A8R32MVP Deluxe.

 

Summary of Performance and Driver Information Reported by ethtool

4.5-Beta64.5-Beta84.5-Beta10

NIC1000Mb/sDriver/Vs1000Mb/sDriver/Vs1000Mb/sDriver/Vs

Marvell 88E8053 PCIeY*/Nsky2/1.22Y**/Nsky2/1.23Nosky2/1.23

Marvell 888001 PCINoskyge/1.13Yesskyge/1.13Yesskyge/1.13

 

  *Only with the secondary NIC disabled will primary NIC negotiate a 1000Mb/s connection with beta 6.

**Only with both NICs enabled will the primary NIC negotiate a 1000Mb/s connection with beta 8.

    Unable to obtain gigabit connection with primary NIC under any circumstances with beta 10.

 

Secondary PCI NIC obtains gigabit connection with betas 8 & 10 but does not support WOL.

Breakthrough, I guess. I now have gigabit working with beta 11. Well, with six out of eight startups anyway, and 75% is a lot better than 0%.

 

The change I made, oddly enough, was to use a different USB flash drive. I recalled that the only time I had problems with beta 6 negotiating a gibabit connection was when I had the flash drive plugged into a port on a USB controller that shared an IRQ with the primary LAN port. So I thought perhaps the flash drive was marginal (it's an old 512MB CF card in a card reader). Apparently there was something to that theory, since the NIC appears to be working significantly better with the new flash drive installed. Why, exactly, I can only guess. Maybe someone with more knowledge will have a theory.

 

Possibly there are several factors at play causing all this weirdness: marginal chip or circuit on my motherboard, combined with marginal drivers, and who knows what else, perhaps something with the later linux kernels.

 

One other interesting change is that now when I tried disabling the secondary NIC, the primary NIC wouldn't work at all (wouldn't negotiate any connection), at least the one time I tried it. I never saw that behavior with beta 6 or 8. In fact with beta 6 that was the most reliable configuration.

 

I guess now I'll see if when the NIC negotiates 1000Mb/s, as it does most of time, if it will maintain it over time through S3 sleep and wake-up cycles. Maybe it will be stable enough to live with.

 

Thanks to everyone offering feedback so far. If any of this makes sense to you I'd appreciate hearing your thoughts.

 

 

[limetech]

Ok, I remember this now & determining that this is a Windows problem.  To make it happen start out with all drives spun-down.  Now start a movie streaming off one of the drives.  There will be a pause of course until the drive spins up.

 

Now while movie is playing, start accessing directories on other disks via same computer & you will see movie pause while other disk spins up.  This is Windows starving it's own media read stream waiting for another directory access stream to finish.  If you go and access directories via another computer, that computer will pause waiting for the directory listings (waiting for disk to spin up) but the movie on the first computer still plays fine.

 

I think the reason this appears as just a momentary glitch when playing music is because the music player buffers enough data that it just starts to empty as the spin up nears completion.

 

Tom,

 

I see this issue on my Mac Mini connected to unRAID via SMB (Using XBMC/Plex). There is the obvious pause when you start playing a media file from a disk that is spun down (which is understandable). While playing a media file, if I'm browsing directories on unRAID or access another disk that has to be spun up, my media will pause for a few seconds (sometimes it's not even a second, sometimes it's 1-2 seconds).

 

Just letting you know.

 

But if while media is playing on the Mac Mini, if you browse directories or access another disk that has to be spun up, on another computer, does the Mac Mini media playback pause/stutter?

[brainbone]

Does anyone know if the paused I/O on drive spin-up issue (where, if you were playing a video from a spinning disk, that video stream would halt if a file on a spun-down disk was accessed, and resume after the disk completely spun up) is fixed in this release?

 

 

Do you have a syslog from when this would happen?

 

No, but it was apparently a long standing issue. (see: http://lime-technology.com/forum/index.php?topic=3237.0 )

 

That said, 4.5-beta11 (and possibly 4.5-beta8, though I didn't try it) seems to have resolved this issue that has haunted me since I started using unRAID.  I haven't done extensive testing since switching to 4.5-beta11, but it looks like I may not need to keep my disks spinning 24/7 any longer! Woohoo!

 

 

Ok, I remember this now & determining that this is a Windows problem.  To make it happen start out with all drives spun-down.  Now start a movie streaming off one of the drives.  There will be a pause of course until the drive spins up.

 

Now while movie is playing, start accessing directories on other disks via same computer & you will see movie pause while other disk spins up.  This is Windows starving it's own media read stream waiting for another directory access stream to finish.  If you go and access directories via another computer, that computer will pause waiting for the directory listings (waiting for disk to spin up) but the movie on the first computer still plays fine.

 

I think the reason this appears as just a momentary glitch when playing music is because the music player buffers enough data that it just starts to empty as the spin up nears completion.

 

In the past, I was able to repeat the issue by streaming video/music on one PC, and then trying to access a file on a spun down unraid drive from a different PC.  I've been trying to repeat the problem with 4.5-beta11, and haven't been able to yet.

 

[WeeboTech]

The change I made, oddly enough, was to use a different USB flash drive. I recalled that the only time I had problems with beta 6 negotiating a gibabit connection was when I had the flash drive plugged into a port on a USB controller that shared an IRQ with the primary LAN port. So I thought perhaps the flash drive was marginal (it's an old 512MB CF card in a card reader). Apparently there was something to that theory, since the NIC appears to be working significantly better with the new flash drive installed. Why, exactly, I can only guess. Maybe someone with more knowledge will have a theory.

 

Possibly there are several factors at play causing all this weirdness: marginal chip or circuit on my motherboard, combined with marginal drivers, and who knows what else, perhaps something with the later linux kernels.

 

Interesting, could it be an issue with IRQ routing/sharing. Perhaps review output of 

cat /proc/interrupts

and/or move the USB flash adapter around to different ports.

 

Odd situation.

[DMAddict]

4.5-beta11 still not any faster for me with regards to uploading. Only getting 14 MB/s to the cache. unRAID 4.4.2 gives me 38+ MB/s.  

 

That should not be the case - please try -beta11 again, do a transfer, and then post your system log.

 

http://pastebin.com/fc7da351 I'm not sure if this is all the info necessary because it didn't seem to change after I uploaded a new file to the array. I pulled the Syslog from unmenu. The file was 2.65 GB and it took 3:12 to up it to the array.

[Romir]

 

But if while media is playing on the Mac Mini, if you browse directories or access another disk that has to be spun up, on another computer, does the Mac Mini media playback pause/stutter?

 

HD playback on a XBMC box was smooth while spinning up and copying files on a couple disks from another pc. Beta11.

[cwr]

Ok, I remember this now & determining that this is a Windows problem.  To make it happen start out with all drives spun-down.  Now start a movie streaming off one of the drives.  There will be a pause of course until the drive spins up.

 

Now while movie is playing, start accessing directories on other disks via same computer & you will see movie pause while other disk spins up.  This is Windows starving it's own media read stream waiting for another directory access stream to finish.  If you go and access directories via another computer, that computer will pause waiting for the directory listings (waiting for disk to spin up) but the movie on the first computer still plays fine.

 

I think the reason this appears as just a momentary glitch when playing music is because the music player buffers enough data that it just starts to empty as the spin up nears completion.

 

Tom,

 

I see this issue on my Mac Mini connected to unRAID via SMB (Using XBMC/Plex). There is the obvious pause when you start playing a media file from a disk that is spun down (which is understandable). While playing a media file, if I'm browsing directories on unRAID or access another disk that has to be spun up, my media will pause for a few seconds (sometimes it's not even a second, sometimes it's 1-2 seconds).

 

Just letting you know.

 

But if while media is playing on the Mac Mini, if you browse directories or access another disk that has to be spun up, on another computer, does the Mac Mini media playback pause/stutter?

 

i see this all the time, using 2 different mac minis and and PC running OSX. if drive-A is spun down, and you're watching a movie from drive-B, if you do something to drive-A causing it to spin up, then you'll see a pause for about 1-2 seconds while the drive is spinning up

[limetech]

Ok, I remember this now & determining that this is a Windows problem.  To make it happen start out with all drives spun-down.  Now start a movie streaming off one of the drives.  There will be a pause of course until the drive spins up.

 

Now while movie is playing, start accessing directories on other disks via same computer & you will see movie pause while other disk spins up.  This is Windows starving it's own media read stream waiting for another directory access stream to finish.  If you go and access directories via another computer, that computer will pause waiting for the directory listings (waiting for disk to spin up) but the movie on the first computer still plays fine.

 

I think the reason this appears as just a momentary glitch when playing music is because the music player buffers enough data that it just starts to empty as the spin up nears completion.

 

Tom,

 

I see this issue on my Mac Mini connected to unRAID via SMB (Using XBMC/Plex). There is the obvious pause when you start playing a media file from a disk that is spun down (which is understandable). While playing a media file, if I'm browsing directories on unRAID or access another disk that has to be spun up, my media will pause for a few seconds (sometimes it's not even a second, sometimes it's 1-2 seconds).

 

Just letting you know.

 

But if while media is playing on the Mac Mini, if you browse directories or access another disk that has to be spun up, on another computer, does the Mac Mini media playback pause/stutter?

 

i see this all the time, using 2 different mac minis and and PC running OSX. if drive-A is spun down, and you're watching a movie from drive-B, if you do something to drive-A causing it to spin up, then you'll see a pause for about 1-2 seconds while the drive is spinning up

 

But does it still do this using -beta11 ?

[cwr]

on beta11, yeah

 

i tried a couple of times to make sure, but it looks like it's still doing it. nothing in the syslog either. even the webui seems to take a few seconds to respond while a drive is spinning up, although that's not really an issue i guess

[Jon]

Ok, I remember this now & determining that this is a Windows problem.  To make it happen start out with all drives spun-down.  Now start a movie streaming off one of the drives.  There will be a pause of course until the drive spins up.

 

Now while movie is playing, start accessing directories on other disks via same computer & you will see movie pause while other disk spins up.  This is Windows starving it's own media read stream waiting for another directory access stream to finish.  If you go and access directories via another computer, that computer will pause waiting for the directory listings (waiting for disk to spin up) but the movie on the first computer still plays fine.

 

I think the reason this appears as just a momentary glitch when playing music is because the music player buffers enough data that it just starts to empty as the spin up nears completion.

 

Tom,

 

I see this issue on my Mac Mini connected to unRAID via SMB (Using XBMC/Plex). There is the obvious pause when you start playing a media file from a disk that is spun down (which is understandable). While playing a media file, if I'm browsing directories on unRAID or access another disk that has to be spun up, my media will pause for a few seconds (sometimes it's not even a second, sometimes it's 1-2 seconds).

 

Just letting you know.

 

But if while media is playing on the Mac Mini, if you browse directories or access another disk that has to be spun up, on another computer, does the Mac Mini media playback pause/stutter?

 

i see this all the time, using 2 different mac minis and and PC running OSX. if drive-A is spun down, and you're watching a movie from drive-B, if you do something to drive-A causing it to spin up, then you'll see a pause for about 1-2 seconds while the drive is spinning up

 

But does it still do this using -beta11 ?

 

Same here and has been bad enough for me to consider going back to a Windows-based server when and if I eventually exceed my capacity.

 

I have 5 Windows Media Centers on my network and any time I start a movie/TV show I will get 2-3 stutters throughout whenever one of the other Media Centers causes additional disks to spin up due to directory access.  The audio continues unaffected, but the video pauses for about 2 seconds and then fast forwards to catch up with the audio...very different from typical audio/video stutter.  I can repeat it any time I wish with any version of unRAID and it appears to be getting worse the more data my unRAID server stores.  No indication in my syslog either.  If I leave all of my disks spinning 24/7, I never encounter that problem, but I don't call that much of a solution considering the disk spin down is a power saving feature that I like to use.

 

I would happily trade the new write performance gains to see this stutter go away.  It's really the only problem I have ever had with unRAID.

[WeeboTech]

If I leave all of my disks spinning 24/7, I never encounter that problem, but I don't call that much of a solution considering the disk spin down is a power saving feature that I like to use.

 

I would happily trade the new write performance gains to see this stutter go away.  It's really the only problem I have ever had with unRAID.

 

Does this happen with direct disk shares or only user shares?

[Jon]

I've never tried with direct disks as I export read-only user shares to all my Media Centers to avoid the "accidental delete".

 

I'll give it a try using the disk share in the next day or two to see if it is specific to user shares and report back here.

[rsd]

Same here and has been bad enough for me to consider going back to a Windows-based server when and if I eventually exceed my capacity.

 

I have 5 Windows Media Centers on my network and any time I start a movie/TV show I will get 2-3 stutters throughout whenever one of the other Media Centers causes additional disks to spin up due to directory access.  The audio continues unaffected, but the video pauses for about 2 seconds and then fast forwards to catch up with the audio...very different from typical audio/video stutter.  I can repeat it any time I wish with any version of unRAID and it appears to be getting worse the more data my unRAID server stores.  No indication in my syslog either.  If I leave all of my disks spinning 24/7, I never encounter that problem, but I don't call that much of a solution considering the disk spin down is a power saving feature that I like to use.

 

I would happily trade the new write performance gains to see this stutter go away.  It's really the only problem I have ever had with unRAID.

 

Exact same issue here, and this is my only true gripe with Unraid. Write performance is nice, but this is extremely annoying. Would love it Tom, if you would be able to investigate it further.

[brainbone]

Same here and has been bad enough for me to consider going back to a Windows-based server when and if I eventually exceed my capacity.

 

I have 5 Windows Media Centers on my network and any time I start a movie/TV show I will get 2-3 stutters throughout whenever one of the other Media Centers causes additional disks to spin up due to directory access.  The audio continues unaffected, but the video pauses for about 2 seconds and then fast forwards to catch up with the audio...very different from typical audio/video stutter.  I can repeat it any time I wish with any version of unRAID and it appears to be getting worse the more data my unRAID server stores.  No indication in my syslog either.  If I leave all of my disks spinning 24/7, I never encounter that problem, but I don't call that much of a solution considering the disk spin down is a power saving feature that I like to use.

 

I would happily trade the new write performance gains to see this stutter go away.  It's really the only problem I have ever had with unRAID.

 

Exact same issue here, and this is my only true gripe with Unraid. Write performance is nice, but this is extremely annoying. Would love it Tom, if you would be able to investigate it further.

 

Looks like I'm still having the issue as well.  It does seems to be happening less, especially when streaming lower bit rate files from unraid, but this is likely to do with client not having its local buffer exhausted before the spin-up completes.  Perhaps some of the optimizations in beta8-beta11 allow the client buffer to "recharge" faster, masking the issue more than was possible in the past -- or perhaps it was all placebo.

 

This is a serious issue for me.  One of the more alluring features of unRAID was the possibility of not having all drives spun up (saving power, wear on HDs, etc.) but with this issue, the spin down feature is basically useless in my situation.

 

[WeeboTech]

Looks like I'm still having the issue as well.  It does seems to be happening less, especially when streaming lower bit rate files from unraid, but this is likely to do with client not having its local buffer exhausted before the spin-up completes.  Perhaps some of the optimizations in beta8-beta11 allow the client buffer to "recharge" faster, masking the issue more than was possible in the past -- or perhaps it was all placebo.

 

This is a serious issue for me.  One of the more alluring features of unRAID was the possibility of not having all drives spun up (saving power, wear on HDs, etc.) but with this issue, the spin down feature is basically useless in my situation.

 

maybe this could be helped with some form of re-ahead optimization and/or extra ram.

I think part of what is needed is determination if it's a shfs or kernel issue.