An Improved unRAID web-interface, extensible, and easy to install

[Eldron]

Forgive me if I have missed the solution to this.

 

I have install unmenu on 4.4.2 and 4.5 beta 11.  I get the same result on each.  The top three boxes Array Status, Array Disk Status and UNRAID drive show.  However, anything below that is just an IE or Firefox page cannot be displayed error.  None of the links work either.  Am I missing a setting I was supposed to change?

What OS are you running?  PC? or MAC OS?

Where did you get the unMENU files from?

 

How did you install them?

 

What happens when you click on a link that does not work?   

 

 

Joe L.

 

 

 

I think I may have found a problem.  I see that  "smartctl: error while loading shared libraries: libstdc++.so.6: cannot open shared object file: No such file or directory" at the console.  I found some references to installing it through the package manager, but none of the subpages work.  I get page cannot be displayed.

 

I am using Windows 7 x64 and Server 2003 x32 with IE8 and the latest Firefox.  I downloaded the files from the Google Code link.  I followed the instructions in this post.  The files were unzipped and placed in the /boot/unmenu folder.  I start unmenu by going to /boot/unmenu and entering uu,

 

Thanks for you help on this.

 

 

[Joe L.]

Forgive me if I have missed the solution to this.

 

I have install unmenu on 4.4.2 and 4.5 beta 11.  I get the same result on each.  The top three boxes Array Status, Array Disk Status and UNRAID drive show.  However, anything below that is just an IE or Firefox page cannot be displayed error.  None of the links work either.  Am I missing a setting I was supposed to change?

What OS are you running?  PC? or MAC OS?

Where did you get the unMENU files from?

 

How did you install them?

 

What happens when you click on a link that does not work?   

 

 

Joe L.

 

 

 

I think I may have found a problem.  I see that  "smartctl: error while loading shared libraries: libstdc++.so.6: cannot open shared object file: No such file or directory" at the console.  I found some references to installing it through the package manager, but none of the subpages work.  I get page cannot be displayed.

I've never used Windows 7, or Server 2003, or even IE8... but I do use IE5/6/7 and Firefox on Vista, Win2000, and WinXP.

I am using Windows 7 x64 and Server 2003 x32 with IE8 and the latest Firefox.  I downloaded the files from the Google Code link.  I followed the instructions in this post.  The files were unzipped and placed in the /boot/unmenu folder.  I start unmenu by going to /boot/unmenu and entering uu,

 

Thanks for you help on this.

The shared library would prevent you from seeing disk temperatures, and from running smart reports, but it would not affect anything else.

For the links not to work would indicate you do not have a name-server on your LAN.  (or are you getting to the server via IP address and not server name?)

 

How are you connected to your server?  (router? switch?)

 

Joe L.

[Eldron]

Interesting, and accurate, observations.  Unmenu works perfectly on the same subnet as the UNraid box.  It is possible some crazy stuff is happening on the switches here at work. Excellent work, and thanks for the help!

 

 

[crankbearing]

 

Joe,

 

I am testing this on my test machine with a newly loaded 4.5 final. My current Unraid Servers are at 4.33. I am thinking of upgrading anyhow. I noticed in browsing through the options of Unmenu, that you use the tool to set read aheads.

 

I have always had the error on my actual unraid media server after it runs the startup and goes multiuser I get cannot find /md* and I know that I put this in my go script from the beginning try to set read ahead to 2048. If I tried to use this tool to add the 2048 option to my script would it overwrite the current lines in my go script where the error is and actually set it right or just add to the script at the bottom causing me more grief.

 

Great Utility looks very promising Awesome Work. I have been away a while, you guys have made great headway.

 

Just added a 2TB Parity drive to the system yesterday, system is at least 5-6 years old now 14 drives and parity-sync on 2TB in under 18hours at 66Kb/sec - It started out much slower though at like 14Kb/sec.

 

Thanks,

 

Dave in Ottawa.

 

 

[Joe L.]

 

Joe,

 

I am testing this on my test machine with a newly loaded 4.5 final. My current Unraid Servers are at 4.33. I am thinking of upgrading anyhow. I noticed in browsing through the options of Unmenu, that you use the tool to set read aheads.

 

I have always had the error on my actual unraid media server after it runs the startup and goes multiuser I get cannot find /md* and I know that I put this in my go script from the beginning try to set read ahead to 2048. If I tried to use this tool to add the 2048 option to my script would it overwrite the current lines in my go script where the error is and actually set it right or just add to the script at the bottom causing me more grief.

The Set-Readahead button makes no changes to your "go" script.  It is simply a short script to loop and set all the re-ahead values when you press it.  The next time you reboot, the values will all be back to their original values.

 

The odds are that you did not allow any time for the array to come on-line before you attempted to set read ahead values, or, you did not give the full path of

/dev/md*

 

There is some good news however, as a result of my discovery of improved performance Tom @ lime-technology did some tests and as a result the delivered re-ahead value is now set to 1024.  This value seemed to give Tom the best overall performance in his tests.  As a result, there is no longer any need t set the read-ahead.

 

Now, if you want to experiment, the unMENU buttons let you play with different values and see what works best in your system.

 

Great Utility looks very promising Awesome Work. I have been away a while, you guys have made great headway.

 

Just added a 2TB Parity drive to the system yesterday, system is at least 5-6 years old now 14 drives and parity-sync on 2TB in under 18hours at 66Kb/sec - It started out much slower though at like 14Kb/sec.

Thanks,

Dave in Ottawa.

Sounds good.  Have fun... lots of changes in the past 5 years.

[Biggy2872]

I want to report a small bug and fix...  Currently the way the package manager is coded and the included (?!?) v1.2 rc.local_startup package file are written, it produces an unexpected error...

 

There is no file to download and therfore no md5 to check.

 

When you load up the package manager on a system without the rc.d TPBA structure, the following error is reported "(MD5 of existing downloaded file NOT matched - download may be corrupted.)" and the button to install the package instead asks if I want to re-download the file. This is despite the fact that there is no file to download and there is no md5 specified in the config file AND the package manager recognizes the fact that it does not need to download a file (message beside the package name says "Package not yet installed (no download needed)"). 

 

I devised one possible solution.

 

990-unmenu-wget.awk

Starts at line 334

# package is not installed yet, but is downloaded
           if ( package_url[i] == "none" ) {
             theHTML = theHTML "<td><font color=\"purple\">Package not yet installed (no download needed)</font></td>"
             nomd5 = "true"
           } else {
             theHTML = theHTML "<td><font color=\"purple\">Package downloaded, but not yet installed</font></td>"
           }
           if ( nomd5 == "true" || VerifyMD5( PACKAGE_DIRECTORY "/" package_file[i], package_md5[i] ) == "OK" ) {
             theHTML = theHTML "<td><input type=submit name=\"manual_install-" package_file[i] "\" value=\"Install " package_file[i] "\"</td>"
             nomd5 = ""
           } else {
             theHTML = theHTML "<td><b><font color=\"red\"> (MD5 of existing downloaded file NOT matched - download may be corrupted.)</b></font>"
             theHTML = theHTML "<input type=submit name=\"download-" package_file[i] "\" value=\"Re-Download " package_file[i] "\"</td>"
           }

I added line 337:

nomd5 = "true"

and added the condition to line 341(previously 340):

if ( nomd5 == "true" || VerifyMD5( PACKAGE_DIRECTORY ...

and finally I added line 343:

nomd5 = ""

 

There are other ways to fix this problem and Joe L. will probably come up with a better one. But this will work in the interim.

 

Cheers,

Matt

[Joe L.]

I want to report a small bug and fix...  Currently the way the package manager is coded and the included (?!?) v1.2 rc.local_startup package file are written, it produces an unexpected error...

 

There is no file to download and therfore no md5 to check.

I remember starting the coding of logic that could install a package without a download of a file, but I don't ever remember implementing it.  (or at least, I started on it, but don't ever remember deploying a package that needed it, so it never was completed.

 

I'll have to try your "fix"  as it might be very useful.  I eventually wanted to get back to the package manager and make some improvements, but never found the time to do so.  I wanted to make the download "url" a field that can be updated per package, since the download link constantly seems to go out-of-date.  It would make it easier to post a list of working download links rather than have to re-deploy a new .conf file.

There are other ways to fix this problem and Joe L. will probably come up with a better one. But this will work in the interim.

 

Cheers,

Matt

I'll try yours first...    There is a small bug in how it handled the "version" number matching, as it did not properly clear a prior value, but that is un-related to what you found.

 

You can fix that by adding one line, shown in blue here, to the PackageVersionTest function

function PackageVersionTest( theTest ) {

  verString=""

  OLD_RS=RS

  RS="\n"

  theTest | getline verString

  close(theTest);

  RS=OLD_RS

  return verString;

}

 

 

Thanks for the feedback, and the proposed fix.  Both are appreciated.  I expect the whole unMENU package manager concept will evolve as unRAID 5.0 evolves.  The package-manager certainly has allowed us to experiment and learn what works, and what does not...

 

Joe L.

[abeta]

I just installed unmenu 1.2 and it seems to be working. I was able to edit the drives information to carry some information for myself.

 

I went into "Package Manager" and selected install "apcupsd - A daemon for controlling APC UPS devices". There was an option to download it, which I did and then it presented me a button for install which I did.

 

It returns this "error" message: Installed, but version is different.

Current version='' expected '3.14.3'

 

There's also a button again that says "Install apcupsd-3.14.3-i486-1kjz.tgz"?

 

I've tried installing it a couple of times and it always puts me back to the same place. Any assistance would be appreciated.

 

Here's the install messages:

apcupsd-3.14.3-i486-1kjz.tgz installation:

 

Verifying package apcupsd-3.14.3-i486-1kjz.tgz.

Installing package apcupsd-3.14.3-i486-1kjz.tgz:

PACKAGE DESCRIPTION:

# apcupsd - A daemon for controlling APC UPS devices

#

# Apcupsd can be used for power management and controlling most of

# APC's UPS models on Unix and Windows machines.  Apcupsd works with

# most of APC's Smart-UPS models as well as most simple signalling

# models such as Back-UPS and BackUPS-Office.

#

# Requires: gd, libjpeg, libpng, ncurses

# Homepage: http://www.apcupsd.org/

#

# Packaged by Ken Zalewski ([email protected])

Executing install script for apcupsd-3.14.3-i486-1kjz.tgz.

Leaving file etc/apcupsd/apcupsd.conf in tact

Package apcupsd-3.14.3-i486-1kjz.tgz installed.

 

 

[purko]

There is some good news however, as a result of my discovery of improved performance Tom @ lime-technology did some tests and as a result the delivered re-ahead value is now set to 1024.  This value seemed to give Tom the best overall performance in his tests.  As a result, there is no longer any need t set the read-ahead.

 

After numerous tests, I can confitm that the default re-ahead value of 1024 gives the best overall performance on my unRAID servers.

 

Purko

 

 

[Joe L.]

I just installed unmenu 1.2 and it seems to be working. I was able to edit the drives information to carry some information for myself.

 

I went into "Package Manager" and selected install "apcupsd - A daemon for controlling APC UPS devices". There was an option to download it, which I did and then it presented me a button for install which I did.

 

It returns this "error" message: Installed, but version is different.

Current version='' expected '3.14.3'

 

There's also a button again that says "Install apcupsd-3.14.3-i486-1kjz.tgz"?

 

I've tried installing it a couple of times and it always puts me back to the same place. Any assistance would be appreciated.

 

Here's the install messages:

apcupsd-3.14.3-i486-1kjz.tgz installation:

 

Verifying package apcupsd-3.14.3-i486-1kjz.tgz.

Installing package apcupsd-3.14.3-i486-1kjz.tgz:

PACKAGE DESCRIPTION:

# apcupsd - A daemon for controlling APC UPS devices

#

# Apcupsd can be used for power management and controlling most of

# APC's UPS models on Unix and Windows machines.  Apcupsd works with

# most of APC's Smart-UPS models as well as most simple signalling

# models such as Back-UPS and BackUPS-Office.

#

# Requires: gd, libjpeg, libpng, ncurses

# Homepage: http://www.apcupsd.org/

#

# Packaged by Ken Zalewski ([email protected])

Executing install script for apcupsd-3.14.3-i486-1kjz.tgz.

Leaving file etc/apcupsd/apcupsd.conf in tact

Package apcupsd-3.14.3-i486-1kjz.tgz installed.

 

 

If you do not have an APC UPS attached to your server, the apcupsd program will not start.  If it does not start, it will not return a version string, that is why it is returning a blank.

 

Basically you need to attach your UPS and then press the "Install" button.

 

Joe L.

[abeta]

If you do not have an APC UPS attached to your server, the apcupsd program will not start.  If it does not start, it will not return a version string, that is why it is returning a blank.

 

Basically you need to attach your UPS and then press the "Install" button.

 

Joe L.

 

Doh! My apologies! I was still running a parity check while I was fiddling. Works much better after I moved it to the UPS. Thx!

[jupilerman]

I upgraded recently to 4.5 final and now, the System Info menu reports are incomplete :

from cpu info : cpu are ok but there is no sensors information :

Sensor info (from /usr/bin/sensors) : No sensors found! Make sure you loaded all the kernel drivers you need.

from list devices :

(from /usr/bin/lsdev) sh: /usr/bin/lsdev: /usr/bin/perl: bad interpreter: No such file or directory

 

Those were correct with 4.4.2 Is there a missing library in 4.5 ? Does unmenu need to be upgrade ?

[Joe L.]

I upgraded recently to 4.5 final and now, the System Info menu reports are incomplete :

from cpu info : cpu are ok but there is no sensors information :

Sensor info (from /usr/bin/sensors) : No sensors found! Make sure you loaded all the kernel drivers you need.

from list devices :

(from /usr/bin/lsdev) sh: /usr/bin/lsdev: /usr/bin/perl: bad interpreter: No such file or directory

 

Those were correct with 4.4.2 Is there a missing library in 4.5 ? Or should unmenu be upgraded ?

unRAID has never had "perl"  (I don't remember if it had "sensors" data... perhaps unRAID used to, but Tom accidentally left it out of 4.5.  I don't ever remember it working on my server's motherboard)

 

all the buttons are doing is showing you the output of the linux commands.    Get them to work at the command line, and unMENU will work too.

[abeta]

I have a question regarding the output on the myMain screen. I have 5 drives (2 Seagate and 3 WD). I've noticed that when the Seagate drives spindown, they show a temperature of *, but when the WD drives spindown they show an actual temp like 31C or 32C, but when I check the unRaid Main screen they show as *. At first I thought it was the last reported temp before it spun down but I've seen the temps drop over time AFTER it has spun down?

 

Why the different output between the Seagate and WD drives on the myMain screen? Is it just coincidence that I see different output by vendor?

 

It isn't a problem, just something bugging my curiosity 

[SSD]

I have a question regarding the output on the myMain screen. I have 5 drives (2 Seagate and 3 WD). I've noticed that when the Seagate drives spindown, they show a temperature of *, but when the WD drives spindown they show an actual temp like 31C or 32C, but when I check the unRaid Main screen they show as *. At first I thought it was the last reported temp before it spun down but I've seen the temps drop over time AFTER it has spun down?

 

Why the different output between the Seagate and WD drives on the myMain screen? Is it just coincidence that I see different output by vendor?

 

It isn't a problem, just something bugging my curiosity 

 

Some drives automatically spin up when you ask them to report their temperature reading, while others are able to report this information without waking the drive.

 

I wrote myMain with some smarts to detect which drives (basically WD drives but can be user over-ridden as well) can report temperature any time.

 

unRAID mixes the state of the drive (the asterisk) with the temperature of the drive using a single column.  myMain has two columns for these two pieces of information.

 

[kapperz]

Dumb question, what command is run to spin down a drive immediately? Just like the button on the disk management screen of unMenu (I did search around first). Thanks

[Joe L.]

Dumb question, what command is run to spin down a drive immediately? Just like the button on the disk management screen of unMenu (I did search around first). Thanks

That depends on the version of unRAID you are running.

[kapperz]

Using 4.5 Final

[Joe L.]

Using 4.5 Final

With 4.5 final you must use the spinup and spindown commands that were newly added in the last of the 4.5beta series.  If you do not, then unRAID's processes will spin up/down a drive as it thinks appropriate, since it will not be aware you spun them up using raw device access, or down using hdparm.

 

The command to spin up a drive in the array is

/root/mdcmd spinup X

where "X" is a number from 0 through 19 corresponding to parity(0) and disk1 through 19

 

the corresponding spindown command is

/root/mdcmd spindown X 

 

The buttons for spinup/spindown on the disk management screen on unMENU for all but the very latest version will not be correctly recognized by unRAID and it will work against you when it thinks it is time to spin them down.

 

An updated set of plug-ins for unMENU's array management and disk management were posted yesterday in this thread:

http://lime-technology.com/forum/index.php?topic=4993.msg46057#msg46057

 

They are described in the "About" page as:

08-unmenu-array_mgmt.awk: 1.2 - modified spin-up/spin down to use commands available as of 4.5 unRAID  - Joe L.
09-unmenu-disk_mgmt.awk: 1.4 - use spinup/spindown commands now available in 4.5 unRAID Joe L.

Joe L.

[kapperz]

Thanks Joe for the quick reply. I updated unMenu as you suggested. I was actually trying to find the command to use on a USB hard drive I have attached to the system. I assume the /root/mdcmd spinup X is not intended for non-array drives.

 

I have one drive installed (attached to one of the raid controllers), but not apart of the array. And another drive (backup) using a HDD dock attached using USB. The USB drive is always spinning. I was trying to find the command to spin down the these drives (like unMenu) and couldn't locate the command to do this

 

(I know it sounds silly at first, but the two drives are Truecrypt partitions and I purposely don't want them part of the array)

[Joe L.]

Thanks Joe for the quick reply. I updated unMenu as you suggested. I was actually trying to find the command to use on a USB hard drive I have attached to the system. I assume the /root/mdcmd spinup X is not intended for non-array drives.

 

I have one drive installed (attached to one of the raid controllers), but not apart of the array. And another drive (backup) using a HDD dock attached using USB. The USB drive is always spinning. I was trying to find the command to spin down the these drives (like unMenu) and couldn't locate the command to do this

 

(I know it sounds silly at first, but the two drives are Truecrypt partitions and I purposely don't want them part of the array)

Many USB interfaces do not support spin-up and spin-down commands, but to spin down a non-array device, IF it supports it, it is as easy as

hdparm -y /dev/XXX

 

where XXX = device name

[betaman]

An updated set of plug-ins for unMENU's array management and disk management were posted yesterday in this thread:

http://lime-technology.com/forum/index.php?topic=4993.msg46057#msg46057

 

They are described in the "About" page as:

08-unmenu-array_mgmt.awk: 1.2 - modified spin-up/spin down to use commands available as of 4.5 unRAID  - Joe L.
09-unmenu-disk_mgmt.awk: 1.4 - use spinup/spindown commands now available in 4.5 unRAID Joe L.

Joe L.

 

Hi Joe, just wondering if this is a general update for unmenu using 4.5 Final or if it just resolves a specific configuration issue?  It wasn't entirely clear to me after reading through the posts.  Thanks!

[Joe L.]

An updated set of plug-ins for unMENU's array management and disk management were posted yesterday in this thread:

http://lime-technology.com/forum/index.php?topic=4993.msg46057#msg46057

 

They are described in the "About" page as:

08-unmenu-array_mgmt.awk: 1.2 - modified spin-up/spin down to use commands available as of 4.5 unRAID  - Joe L.
09-unmenu-disk_mgmt.awk: 1.4 - use spinup/spindown commands now available in 4.5 unRAID Joe L.

Joe L.

 

Hi Joe, just wondering if this is a general update for unmenu using 4.5 Final or if it just resolves a specific configuration issue?  It wasn't entirely clear to me after reading through the posts.  Thanks!

It is an update needed for using unMENU with 4.5final if you want the spin-up/spin-down buttons to work.

 

I'll be incorporating a full set of updated files at some point as unMENU 1.3, but for now, you can get the fixed versions by simply putting into place the two new files for the array management and disk management plug-ins.

 

Joe L.

[kapperz]

Many USB interfaces do not support spin-up and spin-down commands, but to spin down a non-array device, IF it supports it, it is as easy as

hdparm -y /dev/XXX

 

where XXX = device name

Thanks, and you are right. I can not spin down the HD in the USB dock.

[abeta]

I was looking through some pages on hard disk performance and noticed some tests for performance and was wondering if there is a way to put in one of the disk performance test scenarios in either the mymain performance tab or in the disk management page?

 

Thx!