Dynamix - Web GUI


Recommended Posts

Since the last update of GUI 2.2.2, i have also "Ghosts" in my machine:

 

1) Webinterface not reachable

2) Sleep not working propperly

3) Cant shutdown over Putty - very strange

4) After a big amount of Data transfere the server was not longer responsible >> hard reset over power switch

 

The most strange thing is the sleep - it does not any longer what is set in the GUI - strang :o

 

 

Link to comment
  • Replies 1.3k
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

Posted Images

The Web GUI stops loading for me after about a day or two of being up.

The array seems to be up, and I have apps running fine (Plex, SAB, Sickbeard) but the Web GUI stops loading for some reason.

It's not my cache, I have tried it on multiple machines in my house after clearing my caches.

 

I’ve re-installed it once before, and the problem has yet to go away….

 

I thought it was one of the plugins I was using (system stats, so I disabled it and it was up for about 17 days…Had to reboot, and not I can’t get the Web GUI to stay running for more than a few hours.

 

Anyone have any clue what's going on?

 

Link to syslog: https://drive.google.com/file/d/0BweGVGrblPvgS29nVUswbGFackE/edit?usp=sharing

 

Any idea what could be causing this? I looked through my syslog and there didn't seem to be anything dynamix related in there.  Is there something I can post that will help diagnose this?

Link to comment

The Web GUI stops loading for me after about a day or two of being up.

The array seems to be up, and I have apps running fine (Plex, SAB, Sickbeard) but the Web GUI stops loading for some reason.

It's not my cache, I have tried it on multiple machines in my house after clearing my caches.

 

I’ve re-installed it once before, and the problem has yet to go away….

 

I thought it was one of the plugins I was using (system stats, so I disabled it and it was up for about 17 days…Had to reboot, and not I can’t get the Web GUI to stay running for more than a few hours.

 

Anyone have any clue what's going on?

 

Link to syslog: https://drive.google.com/file/d/0BweGVGrblPvgS29nVUswbGFackE/edit?usp=sharing

 

Any idea what could be causing this? I looked through my syslog and there didn't seem to be anything dynamix related in there.  Is there something I can post that will help diagnose this?

 

Dynamix WebGui has only locked/stopped responding on me once and I was updating The SNAP plugin when my situation happened. All of my shares and other plugins were still functional except - unMenu and SNAP. Had to shutdown cleanly using Command Line. Hasn't stopped working since.. Just thought I would post my situation in which WebGui acted up...

Link to comment

I have no problem with dynamix at all. 2.2.2 is working fine + sleep + apcupsd + shares.

 

Yea, It's strange, it works fine...then all of a sudden the Webgui is unresponsive.

Array, shares and apps are all working great still.

I had the Webgui up for about 3 weeks...one day it just stops..I'll restart it, and sometimes its up for another few days and it will stall again... It's very strange.

Link to comment

is this a good GUI ?

 

how are gui's able to improve unraid other than aesthetically? can you fork a build with scripts and other bits to make your own unraid (post paying for it of course) 

 

seems like tom could benefit from this too..

Link to comment

is this a good GUI ?

 

how are gui's able to improve unraid other than aesthetically? can you fork a build with scripts and other bits to make your own unraid (post paying for it of course) 

 

seems like tom could benefit from this too..

See the github link at the bottom of the first post for some idea of the extra features that can be used with this one.
Link to comment

I have no problem with dynamix at all. 2.2.2 is working fine + sleep + apcupsd + shares.

 

Yea, It's strange, it works fine...then all of a sudden the Webgui is unresponsive.

Array, shares and apps are all working great still.

I had the Webgui up for about 3 weeks...one day it just stops..I'll restart it, and sometimes its up for another few days and it will stall again... It's very strange.

 

Sure it is strange. I restart a month ago but not for a problem, just a socket change in the apc ups (server not master now because sleep 20 hours/day).

 

416341652.249474.jpg.09c93cc9331a265a1cf3cc93994feac6.jpg

Link to comment

I have no problem with dynamix at all. 2.2.2 is working fine + sleep + apcupsd + shares.

 

Yea, It's strange, it works fine...then all of a sudden the Webgui is unresponsive.

Array, shares and apps are all working great still.

I had the Webgui up for about 3 weeks...one day it just stops..I'll restart it, and sometimes its up for another few days and it will stall again... It's very strange.

See here
Link to comment

I had it running on March 8th (as previously posted), then the webUI died on me again (on the 10th, I believe). Same symptoms as the last several people have posted. UI is dead, but the server's running just fine. unMENU, SAB, SB, CP, Subsonic, Shares - everything else is good.

 

This time, I don't see anything flagged as an error in the syslog. (attached)

<snip>I'll restart it<snip>

How do you do that? Are you rebooting, or is there a command line to do it?

 

See here

I looked at that post, and a couple around it and it looks like that's in reference to a complete system hang. Looks like the user there isn't able to access shares, or anything.

syslog-2014-03-12.zip

Link to comment

I have no problem with dynamix at all. 2.2.2 is working fine + sleep + apcupsd + shares.

 

Yea, It's strange, it works fine...then all of a sudden the Webgui is unresponsive.

Array, shares and apps are all working great still.

I had the Webgui up for about 3 weeks...one day it just stops..I'll restart it, and sometimes its up for another few days and it will stall again... It's very strange.

See here

 

Thanks trurl,

 

But I have SMB enabled, and I can't take that out...also it looks like that user may be having a different problem.  He isn't able to access any of his shares, whereas my issue is solely the webGUI.

Link to comment

 

How do you do that? Are you rebooting, or is there a command line to do it?

 

 

I am using the powerdown command.  I highly recommend you install this (You should be able to find it via unmenu).

You simply issue that command, and it will stop all running apps, unmount your array and powerdown safely.

Link to comment

Accessing your shares is a function of smb (if using smb for sharing). Accessing the webGUI is a function of emhttp. You can have one still working without the other.

 

In Linux, if some process needs memory and another process is not currently active, then that inactive process can be killed. So if sabnzbd needs more memory, and you are also streaming a movie in XBMC, you could have a situation where smb is active because XBMC is using it, sabnzbd is asking for more memory, and emhttp is not active because nobody is using the webGUI at that time. emhttp gets killed and so you can't access the webGUI next time you try, but smb didn't get killed so your shares still work. This is just an example. Maybe this specific scenario does not apply to you, but some other might.

 

emhttp getting killed is talked about in lots of threads in the forum. Some might even consider it a bug. There is a way of setting process priority that may prevent this from happening. That is the purpose of the lines in the go script that I linked you to.

 

Since Dynamix installs the powerdown script now, you should be able to execute it from the command line and at least safely shutdown so you can reboot.

Link to comment

 

How do you do that? Are you rebooting, or is there a command line to do it?

 

 

I am using the powerdown command.  I highly recommend you install this (You should be able to find it via unmenu).

You simply issue that command, and it will stop all running apps, unmount your array and powerdown safely.

Thanks, I have that and have used it. My question was in relation to restarting Dynamix when it's died. I was under the impression that you had restarted Dynamix without restarting the server. I can do a shutdown/reboot from unMENU, since that's still responsive.

 

Accessing your shares is a function of smb (if using smb for sharing). Accessing the webGUI is a function of emhttp. You can have one still working without the other.

 

In Linux, if some process needs memory and another process is not currently active, then that inactive process can be killed. So if sabnzbd needs more memory, and you are also streaming a movie in XBMC, you could have a situation where smb is active because XBMC is using it, sabnzbd is asking for more memory, and emhttp is not active because nobody is using the webGUI at that time. emhttp gets killed and so you can't access the webGUI next time you try, but smb didn't get killed so your shares still work. This is just an example. Maybe this specific scenario does not apply to you, but some other might.

 

emhttp getting killed is talked about in lots of threads in the forum. Some might even consider it a bug. There is a way of setting process priority that may prevent this from happening. That is the purpose of the lines in the go script that I linked you to.

 

Since Dynamix installs the powerdown script now, you should be able to execute it from the command line and at least safely shutdown so you can reboot.

 

Got it, I understand much better now, trurl, thanks. Your explanation seems very valid, except...

 

A) unMENU is still responsive. Or does that not rely on emhttp?

 

B) I keep a ton of tabs open in Firefox at all times, including Dynamix. Maybe Firefox puts background tabs in their own threads and they go to sleep, but I usually have the stats page open (with the 4 graphs), and whenever I click on that tab, all the graphs are up to date. By that, I mean that there are lots of spikes in activity, not a straight line from the left edge of the graph where it's last known data point was 3 hours ago, to the newest data point at the right edge of the graph. So, with that in mind, I would think that emhttp would be kept active by Dynamix constantly refreshing the graphs and redrawing the page.

 

Also, the memory usage graph usually shows 7GB of RAM used, .5GB free. Of the 7GB used, about 8-900MB is application space, the other 6+GB is cache. Wouldn't Linux pull from the cache to feed apps before shutting down background tasks? (Working off of logic here, not any specific Linux knowledge.)

Link to comment

Don't fully understand some of the details myself. It may be that low memory is all this matters for this.

 

unMenu is really an awk script pretending to be a webserver. I have usually found that it continues to work after the webGUI doesn't. In fact, there is a user script in unMenu that will let you shutdown safely and I have used that when I have lost the webGUI.

 

You can check for the process with

ps aux | grep emhttp

You should get something like:

root@unSERVER:~# ps aux | grep emhttp
root      6472  0.0  0.0   5088  1056 pts/1    S+   20:24   0:00 grep emhttp
root     12079  0.0  0.1  84640  1896 ?        Sl   Mar08   2:18 /usr/local/sbin/emhttp
root@unSERVER:~#

I doubt Dynamix is to blame for this problem. It is still emhttp under the skin.

Link to comment

OK, very bizarre!

 

I did

ps aux|grep emhttp

It shows /usr/local/sbin/emhttp is running. Just for giggles, I refreshed the tab Dynamix is supposed to be in, and there it is.

 

I have not rebooted. I have run no other commands on the console or via putty.

 

The only thing I did was I had a putty session that's been open for a couple of days. I use mc to moved files from a temp directory on my cache drive into their permanent homes on /mnt/user/... I exited the shell, saw your message, then logged back in and ran the ps command.  The only remotely logical thing is that somehow the telnet session running for a long period of time was interfering with emhttp responding.

 

I'm going to leave it connected until Dynamix stops responding, close it immediately, and see if it responds. Won't make sense, but at least we'll have found some sort of causality and or high level of coincidence.

Link to comment

OK, very bizarre!

 

Yea, that is strange.  I ran ps aux | grep emhttp, and got the following:

 

root@Tower:~# ps aux | grep emhttp
root      2563  0.0  0.0   2452   600 pts/0    S+   07:49   0:00 grep emhttp
root     28174  0.0  0.0   6716  1472 ?        Sl   Mar08   0:51 /usr/local/sbin/emhttp
root@Tower:~# 

 

Still no webGUI.

 

Edit: For what its worth, I haven't had any active Telnet sessions open longer than a few minutes.

Link to comment

Bonienl-

 

Just a thought for a future enhancement...

 

Obviously, something is checking somewhat regularly for updates to WebGUI and the various plugins. Maybe have it change the color of the 'Dynamix' tab in the main menu when it finds that there is something out of date. That's not a tab I was checking very often - since I'd installed the plugins I was interested in, why bother going back? A highlight there would help get people to notice and pick up new fixes.

 

Thanks for all your great work on this. I've been having minor issues here and there, but overall very impressive!

 

EDIT: OK, a minor issue to report. Just updated to WebGUI 2.2.3, and the stats part of the dashboard isn't working. See attached screen shot. I do have System Stats 2.1.2 installed and running, but I've noticed that each time I go to the system stats tab, it restarts from 0 - earlier versions seemed to have continued to run in the background, gathering info, even when that tab wasn't being viewed.

unRAID.Dynamix.StatsMissing.PNG.5e0a159e9614e86271da950231cd9b93.PNG

Link to comment

for me it is the other stats screen actually that stopped working

everything else is looking and working just peachy

as you can see it has been running for 3 days but no info whatsover.. the only thing that does work is setting to real time

system stats is 2.1.2 and webgui is 2.2.3

 

while i'm at it anyway.. could we get the button with changes before we update? that would make quite a bit more sense in my world  :o

 

unraid-dynamix-statsscreen.jpg.a32a766d000e277c552f863ba8e1c2ea.jpg

Link to comment

So, I restarted my array again, and monitored the webGUI for the few days it was up.

It lasted about 2 days and then I couldn't reach it again.

 

One thing I noticed before the WebGUI went down is that the Dashboard showed that my memory usage was running pretty high, even when the disks where spun down. I'm not sure if that is normal behavior or not, but I have 16GB of RAM in this machine (overkill, I know).

 

When I first noticed it, it was sitting around 50%.  Over the next day, it went up to 80% then 97% then back down to 80%.  Then next time I looked I couldn't connect.  However, I could pull it up on unRAID menu and it says I was still at 80%.

 

Freeman - Are you still experience the hanging webGUI?

Link to comment
  • Squid locked this topic
Guest
This topic is now closed to further replies.