unRAID Server Release 6.1.2 Available

[monkeymajik]

Sounds like you have a bad NIC port on your motherboard then.

 

The only way to verify that would be to try one of the PCI NICS.  It looks like monkeymajik may be at the point of "absolutely have to" with a PCI NIC to isolate the issue.

 

Agreed, will power down array now and try.

 

The PCI NICs are a completely different chipset, so I'm fairly sure they will work fine.  I was just hoping I could get the onboard working again with a different driver.

[GFOviedo]

Sounds like you have a bad NIC port on your motherboard then.

 

The only way to verify that would be to try one of the PCI NICS.  It looks like monkeymajik may be at the point of "absolutely have to" with a PCI NIC to isolate the issue.

 

Agreed, will power down array now and try.

 

The PCI NICs are a completely different chipset, so I'm fairly sure they will work fine.  I was just hoping I could get the onboard working again with a different driver.

 

I had a similar issue with my onboard as well.  I ended up with a PCI Intel NIC and fixed the issue.

[monkeymajik]

 

Agreed, will power down array now and try.

 

The PCI NICs are a completely different chipset, so I'm fairly sure they will work fine.  I was just hoping I could get the onboard working again with a different driver.

 

PCI NIC in and working at 1000mbit.  Makes me wary of the condition of the rest of the board 

[uldise]

 

Agreed, will power down array now and try.

 

The PCI NICs are a completely different chipset, so I'm fairly sure they will work fine.  I was just hoping I could get the onboard working again with a different driver.

 

PCI NIC in a working at 1000mbit.  Makes me wary of the condition of the rest of the board 

 

try to boot other OS from livecd(Ubuntu for example) to double check

[monkeymajik]

 

PCI NIC in a working at 1000mbit.  Makes me wary of the condition of the rest of the board 

 

try to boot other OS from livecd(Ubuntu for example) to double check

 

Of course, clever idea - Will try.

[jphipps]

Not sure if this belongs in this thread or general support, but I noticed under 6.1.2 that the log display through the UI seems to just hang and most of them time never displays.  I tested this on both of my upgraded unraid servers and it does it across both.

[trurl]

Not sure if this belongs in this thread or general support, but I noticed under 6.1.2 that the log display through the UI seems to just hang and most of them time never displays.  I tested this on both of my upgraded unraid servers and it does it across both.

On Dashboard, under System Status, where it says "flash : log : docker" what is the percent used for log, the middle number?

[jphipps]

Not sure if this belongs in this thread or general support, but I noticed under 6.1.2 that the log display through the UI seems to just hang and most of them time never displays.  I tested this on both of my upgraded unraid servers and it does it across both.

On Dashboard, under System Status, where it says "flash : log : docker" what is the percent used for log, the middle number?

 

It shows 9% on both servers.

[PhantomII]

Not sure if this belongs in this thread or general support, but I noticed under 6.1.2 that the log display through the UI seems to just hang and most of them time never displays.  I tested this on both of my upgraded unraid servers and it does it across both.

On Dashboard, under System Status, where it says "flash : log : docker" what is the percent used for log, the middle number?

 

I've noticed the same thing, however although I'm not 100% certain, it seems to me I had the same problem on 6.1.1

And it's not most of the time, I've never gotten the logs to work.

 

And for what it's worth, my log percentage on the status page is 22%

[Frank1940]

Not sure if this belongs in this thread or general support, but I noticed under 6.1.2 that the log display through the UI seems to just hang and most of them time never displays.  I tested this on both of my upgraded unraid servers and it does it across both.

On Dashboard, under System Status, where it says "flash : log : docker" what is the percent used for log, the middle number?

 

It shows 9% on both servers.

 

Are you blocking popups?

[jphipps]

I get the popup window, but then it just says "waiting for {ip address}" and never seems to return.  I can see the tail command running in the process list.

 

https://www.dropbox.com/s/7pp8peeq20f5962/Screen%20Shot%202015-09-13%20at%2010.34.17%20AM.png?dl=0

[interwebtech]

verified working here using Chrome v45.0.2454.85 m, on 6.1.2

[aaronwt]

verified working here using Chrome v45.0.2454.85 m, on 6.1.2

 

Yes. Logs are working here in Chrome too in Windows 10. Also in the Edge Browser. And in Internet Explorer.

[gubbgnutten]

Weird update experience for me, but all sorted now. Might try to replicate on my lab server out of curiosity. Or possibly actually search the forum...

 

Step-by-step recap in case anyone is curious or easily amused:

 

[*]Upgraded to 6.1.2 from 6.0 via the plugin, rebooted as usual, and ... nothing! Trusty old server.local wouldn't respond! No web interface! No ping! Gaah!

[*]Plugged in display, greeted by normal Linux console prompt and correct version of unRAID. So far so good. But why can't I reach it over the network?

[*]Logged in as root locally. No password required. Unexpected. Got a bad feeling about this...

[*]OK, the server had an IP address, but was not using LACP bonding but rather eth0. Driver issue or lost config?

[*]Outgoing pings functional. SSH connection to IP address instead of hostname from desktop refused due to changed server keys. Telnet OK. Alright, LAN is there too, just reset?

[*]Let's have a look at the config ... hmm, config is completely empty save for ssh. Could explain symptoms so far at least.

[*]What!? Rest of /boot hierarchy is completely empty too, how did the system even boot!?

[*]Alright, default config means no array, and /boot is weird, so nowhere to store diagnostics or syslog... Maybe use the network? Nah, forget it, I'll just remake the USB stick and restore from backup... Want the server up and running ASAP.

[*]Shut down server (long power button press required, not surprising), plugged in USB stick to regular computer.

[*]What!? The USB stick looks perfectly fine, and reports no file system errors. Let's eject it without changing a thing and plug it into the server again...

[*]Server boots just fine.

 

[mr-hexen]

Weird update experience for me, but all sorted now. Might try to replicate on my lab server out of curiosity. Or possibly actually search the forum...

 

Step-by-step recap in case anyone is curious or easily amused:

 

[*]Upgraded to 6.1.2 from 6.0 via the plugin, rebooted as usual, and ... nothing! Trusty old server.local wouldn't respond! No web interface! No ping! Gaah!

[*]Plugged in display, greeted by normal Linux console prompt and correct version of unRAID. So far so good. But why can't I reach it over the network?

[*]Logged in as root locally. No password required. Unexpected. Got a bad feeling about this...

.....

• Server boots just fine.
  

 

Honestly, I've seen MANY reports similar to this and it's why I've stuck with 6.0.1 for the time being...

 

@Limetech, have you guys looked into reports such as this?

[danioj]

Upgraded Backup and Main Servers (In that order) using WebGUI Plugin method from 6.1.0. Perfect experience, no issues.

 

I am REALLY loving this new method of upgrading versions. @LimeTech AFAIAK this is by far the best feature you have released since 5.x.

 

#HappyChappie

 

EDIT:

 

P.S. Also liking the new Release Notes Format. Easy on the eyes, easy to understand. 

[Mettbrot]

Weird update experience for me, but all sorted now. Might try to replicate on my lab server out of curiosity. Or possibly actually search the forum...

 

Step-by-step recap in case anyone is curious or easily amused:

 

[*]Upgraded to 6.1.2 from 6.0 via the plugin, rebooted as usual, and ... nothing! Trusty old server.local wouldn't respond! No web interface! No ping! Gaah!

[*]Plugged in display, greeted by normal Linux console prompt and correct version of unRAID. So far so good. But why can't I reach it over the network?

[*]Logged in as root locally. No password required. Unexpected. Got a bad feeling about this...

[*]OK, the server had an IP address, but was not using LACP bonding but rather eth0. Driver issue or lost config?

[*]Outgoing pings functional. SSH connection to IP address instead of hostname from desktop refused due to changed server keys. Telnet OK. Alright, LAN is there too, just reset?

[*]Let's have a look at the config ... hmm, config is completely empty save for ssh. Could explain symptoms so far at least.

[*]What!? Rest of /boot hierarchy is completely empty too, how did the system even boot!?

[*]Alright, default config means no array, and /boot is weird, so nowhere to store diagnostics or syslog... Maybe use the network? Nah, forget it, I'll just remake the USB stick and restore from backup... Want the server up and running ASAP.

[*]Shut down server (long power button press required, not surprising), plugged in USB stick to regular computer.

[*]What!? The USB stick looks perfectly fine, and reports no file system errors. Let's eject it without changing a thing and plug it into the server again...

[*]Server boots just fine.

 

Ever since 6.0 I have kind of the same issue: When I update and reboot unraid It hangs at loading the bzroot file. When after 10 minutes of waiting I unplugged the usb (server still running) I briefly see some messages like:

EDD: Error 8000 reading sector 347635

and unraid starts to boot. When finished booting I have the same symptoms as gubbgnutten: nearly empty /boot; empty /mnt no /var/log ect.

 

Checking the device on windows yields no errors and when I try this a few times it will eventually boot into unraid without problems (including every following boot). Very weird.

I checked the usb drive for bad sectors with http://rufus.akeo.ie but it seems to be ok.

[RobJ]

Concerning the /boot issue, I think there may be 2 problems here, creating a race condition in the mounting of /boot.  One is the potential delay introduced by the fsck check of the flash drive (added in 6.1.0), and the other being an early creation of /boot/config, possibly by Dynamix by a 'makedir -p'.  If you look at the screen shot here, you see at first glance a correct /boot with config and logs folders, but NO syslinux and NO files!  The /boot/logs folder is created by the previously run diagnostics command.  So at an early point, Dynamix appears to be saving something to the flash drive, in the config folder, and probably using 'mkdir -p /boot/config/plugins/xyz*'.  There's nothing wrong with the 'mkdir -p' for the last part, but it's a serious error if /boot or /boot/config do not yet exist, and force-creating them compounds the problems.  This results in the symptoms seen by the earlier reports.  So first fix is that there CANNOT be a 'mkdir -p' on /boot or /boot/config.  They must be checked for existence, and either wait (with timeout) or report an error (and log it).

 

The other factor that may be making it more likely is if the dosfsck of the flash drive takes too long, allowing the early wrong /boot mount, which then blocks the correct /boot mount of the flash drive.  Right now, dosfsck output only goes to an attached console.  It would be great if that could be echoed to the syslog too.  And it may be necessary to provide users options to control its running - run on next boot only, don't run, always run [the default].  But more importantly, there needs to be blocking on the completion of dosfsck, before additional commands involving its contents can proceed, most especially the mounting of /boot.

 

Qualifier: the screen shot I linked to looks a lot like the issue here, but it apparently happened in 6.0 too (see earlier posts there), so *may* not be the same issue.  It's extremely unlikely, but it *could* be user error.  But I don't think so, flash drive looks correctly prepared and has booted 6.1.2, so I'm leaning towards a slow mounting flash and too early /boot creation.

[bonienl]

The first occurence where sub-folder may be created under /boot is during the execution of rc.local, this is the folder creation for notifications when they are saved to flash. At that point the flash drive should be mounted already.

 

For those experiencing the issue: how is Store notifications to flash set? And if set to YES, can you change to NO and retest?

 

[gubbgnutten]

The first occurence where sub-folder may be created under /boot is during the execution of rc.local, this is the folder creation for notifications when they are saved to flash. At that point the flash drive should be mounted already.

 

For those experiencing the issue: how is Store notifications to flash set? And if set to YES, can you change to NO and retest?

 

Isn't the Store notifications to flash setting just read from the flash drive? Wouldn't that mean that the drive has to already be properly mounted for the value of the setting to differ from the default value (whatever that may be)? And if the drive is already properly mounted, /boot should be perfectly safe to use?

 

Edit: When is /boot/config/ssh created/populated? Don't recall seeing any other files/folders under /boot in my case, but then again, I didn't look for hidden files...

[aaronwt]

I did a parity check recently and it finished this morning around 2AM. I have my spindown timer set for 3 hours so around 5Am all drives spun down. Around 8:40 Am the drives spun up when I was looking at the share. It's now been over ten minutes since the drives spun up and on the main page it still isn't showing any Drive temps. If I go to the dashboard and click on the individual drives, it will show the temps with the SMART info, but the temps are not showing up in the main page. Is this normal behavior?

 

EDIT: The only drive not spin up was the parity drive. When I hit the "SPin UP" button the parity drive did spin up and then all temps showed up on the main page. But I wouldn't think I should need to hit this button for the temps to show up since the 23 data drives were previously spun up.

[RobJ]

I did a parity check recently and it finished this morning around 2AM. I have my spindown timer set for 3 hours so around 5Am all drives spun down. Around 8:40 Am the drives spun up when I was looking at the share. It's now been over ten minutes since the drives spun up and on the main page it still isn't showing any Drive temps. If I go to the dashboard and click on the individual drives, it will show the temps with the SMART info, but the temps are not showing up in the main page. Is this normal behavior?

 

EDIT: The only drive not spin up was the parity drive. When I hit the "SPin UP" button the parity drive did spin up and then all temps showed up on the main page. But I wouldn't think I should need to hit this button for the temps to show up since the 23 data drives were previously spun up.

Change Tunable (poll_attributes) (Settings -> Disk Settings) to something like 180 (3 minutes) or 120 (2 minutes) or 60 (1 minute).  (See Configuring the Settings -> Disk Settings -> the note about Tunable (poll_attributes)).

[MisterWolfe]

Weird update experience for me, but all sorted now. Might try to replicate on my lab server out of curiosity. Or possibly actually search the forum...

 

Step-by-step recap in case anyone is curious or easily amused:

 

[*]Upgraded to 6.1.2 from 6.0 via the plugin, rebooted as usual, and ... nothing! Trusty old server.local wouldn't respond! No web interface! No ping! Gaah!

[*]Plugged in display, greeted by normal Linux console prompt and correct version of unRAID. So far so good. But why can't I reach it over the network?

[*]Logged in as root locally. No password required. Unexpected. Got a bad feeling about this...

.....

• Server boots just fine.
  

 

Honestly, I've seen MANY reports similar to this and it's why I've stuck with 6.0.1 for the time being...

 

@Limetech, have you guys looked into reports such as this?

 

I had the same issue.  In my case, I was unable to boot from the usb key after the upgrade to 6.1.2.  Inspection of the usb key on a different computer showed all the config was wiped out, along with all plugins.  I ended up starting fresh on a new usb key just in case if was a corruption issue on the drive itself.

 

I will be much more careful with upgrades moving forward, particularly after reading other reports of this happening to other people.

 

 

[gubbgnutten]

I had the same issue.  In my case, I was unable to boot from the usb key after the upgrade to 6.1.2.  Inspection of the usb key on a different computer showed all the config was wiped out, along with all plugins.  I ended up starting fresh on a new usb key just in case if was a corruption issue on the drive itself.

 

I will be much more careful with upgrades moving forward, particularly after reading other reports of this happening to other people.

 

To be fair, that does not really sound like the same issue at all if you read past the first point

 

In any case, I don't think you need to be too concerned about upgrading because of some problematic upgrades, just keep backups.

[RobJ]

I had the same issue.  In my case, I was unable to boot from the usb key after the upgrade to 6.1.2.  Inspection of the usb key on a different computer showed all the config was wiped out, along with all plugins.  I ended up starting fresh on a new usb key just in case if was a corruption issue on the drive itself.

As far as I know (and I could be wrong!), I don't believe any of the earlier reports included any damage to the flash drive, like files lost.  My read has been that in each case the flash drive was fine, before and after, but 'ignored'.  When you checked the flash on a different computer, did you try a Check Disk on it?