Re: [cobalt-users] RaQ4r In Constant Reboot! - URGENT!

["tkirk@xxxxxxxxxxxxxx" <tkirk@xxxxxxxxxxxxxx>]

> Please help me stop my RaQ4r from acting like a Windows box...
> 
> I have a slight, *ha*, problem...  Our one and only RaQ4r is stuck 
> in a
> constant reboot loop..  It seems that it'll boot fine, and 
> everything will
> appear ok.  But upon closer inspection, "cat /proc/mdstat" we see 
> that the
> second drive is being rebuilt by the first.  Ok, fine..  But the 
> time just
> keeps going up and up and up and up..  Hmmm, same thing happening 
> in the web
> admin panel..  Not good...  Do some more research, "ps aux" & 
> "top", and it
> seems that ChiliSoft is eating lots more ram and CPU resources 
> than anything
> else..  Ok, so let's do this, "cd /home/chilias"./stopcaspd", 
> GREAT, now when I "cat /proc/mdstat" and check the web admin
> panel, the time is actually going down and the percentage complete 
> is now
> going up!  Wonderful!
> 
> But that didn't solve the problem completely, apparently, there is 
> ALSOsomething else compounding the situation.  Because within 15 - 
> 20 minutes of
> the RAID finishing it's resync, it magically reboots again..  
> Sometimes it
> doesn't even reboot..  Sometimes it just locks and ALL lights on 
> the front
> and back go dead!  Only the LCD is left lit up, BUT the buttons, 
> "S" and "E"
> and the arrows don't work at all..  I tried holding in all of them 
> (one by
> one), nothing..  Flip switch in back, try again..  Hmmm, starting 
> to sound
> like a Window's box..
> 
> Now, we did notice that when the indicator lights on the front 
> blink, so
> does the LCD screen..  You know, like a power drain is 
> happening...  We did
> get it back and stable by stopping Chilisoft from starting and 
> unpluggingthe 2nd hard drive..  All seems to well again, and the 
> box has so much free
> resources, I just can't believe it...  And when the indicator 
> lights blink,
> the LCD doesn't blink anymore, at least it's not noticeable..
> 
> Now, go with me here..  This is my idea as to what is happening, 
> and maybe
> you guys and gals could help me out...  I think that the CPU is 
> not getting
> enough power (dimming of LCD during activity), which is one reason 
> why it's
> stable again after removing 2nd drive?  Also, and ChiliSoft, which 
> eatssystem resources, could be putting unnecessary strain on the 
> CPU causing it
> to consume more and more power until finally it BUSTS and reboots?
> 
> Now this is just a thought, based on the dimming of the LCD and 
> the fact
> that the drives wouldn't rebuild unless ChiliSoft is shut off..  
> And the
> drives are showing up as ok..  They show a normal response (cat
> "/proc/mdstat") when RAID gets to finish it's rebuild..  But the 
> system is
> still unstable with both drives connected, but unplug one drive (power
> cable) and BAM, stable Linux again..
> 
> I've been in the archives and every where I can go, please tell me 
> someoneknows something..  Or even if I am on the right track...
We have had the exact same thing twice with our one and only RAQ4R :( 
Here is what we have done to correct the situation. It seems the RAID 
rebuild takes awhile. Both times for us it has been about 30mins.

go into control panel and turn asp server off and hit save, if from 
here the remaining minutes steadily go down then let it go until 
complete...if not remaining minutes not going down within say five 
minutes (it does take a little while to register) hit properties in asp 
admin then hit 'stop'. This will stop the asp if it is still running. 
We have actually had to do the second step on both occassions.

For some reason the asp serer does not like to stop from the control 
panel, so we have to use the stop feature in the asp admin section. The 
raid rebuild just keeps restarting if teh asp serer is running. That is 
what was our problem both times.

----------------
Powered by telstra.com