Page 1 of 2 12 LastLast
Results 1 to 20 of 21
  1. Tmillz's Avatar
    Posts
    9 Posts
    Global Posts
    14 Global Posts
       #1  
    I wanted to bring to someones attention that webos 1.4 freezes during restart (at spinning circle) when ez-ip update and lighttpd are configured as described in these posts. How to SSH into Your Palm Pre! | Palm Pre Hacks Blog! and WebOS Internals wiki lighttpd. sorry i cant post links yet..

    I know narrowed the problem down to this because after updating to 1.4 the first time i had this problem, so i doctored my pre, then installed everything one by one back on and after ez-ipupdate and lighttpd i had the same problem again.

    i don't know if its a problem with the start up script

    #!/bin/sh
    /opt/bin/ez-ipupdate -c /opt/etc/ipupdate.conf -q > /dev/null 2>&1 || /bin/true

    or the start up script for lighttpd or the packages themselves. Can someone smarter than me look into it? I would be nice if i didn't have to pull my bat. everytime i choose restart. P.S. Shutdown seems to work fine!

    Thanks
  2. Tmillz's Avatar
    Posts
    9 Posts
    Global Posts
    14 Global Posts
       #2  
    anyone?
  3. #3  
    i am experiencing something similar. after upgrading to webos 1.4, my sprint palm pre can no longer reboot or shutdown fully. the phone freezes while the circle with arrows is displayed on the screen and no amount of waiting (at least up to 30 minutes) seems to let it get through whatever it is stuck on.

    i believe my phone installation is fairly "clean". i have a number of patches and optware but nearly everything was installed via preware. i did an EPR before upgrading to 1.4 just incase AUPT hasn't been fully shaken out yet. beyond the optware packages for ez-ipupdate and lighty i added sudo and php-fcgi, that's about it.

    i am trying to debug this at the shell level but so far have not been able to find clues. i doubt ez-ipupdate is the cause because this is not a daemon, rather it is a script that runs whenever ppp.d changes the ip address. lighttpd is certainly a possibility - my first inclination was that perhaps the shutdown/reboot sequence in rc0.d/rc6.d are failing to stop it. this was partially confirmed when i found that stopping lighttpd via initctl isn't always working (/sbin/stop mobi.optware.lighttpd) - rather initctl would report status that it was waiting, and lighttpd would still be running as per 'ps -ax'. however, manually killing lighttpd and then shutting down the pre yielded the same freeze, so i am now skeptical that this is the problem.

    it would be nice if i could step through a log of what the pre is doing while shutting down to see where it gets stuck, but it seems i am not looking in the right place for this information. did not find the right information in /var/log/messages as one might expect. i saw a number of calls to logger in the /etc/init.d scripts but i can't find any logs in /var/log that contain those strings. does anyone know where the logger strings go or a better way to follow the log of operations involved in shutting down webos?
  4. #4  
    Had a similar problem freezing on restart. Did you try the suggestion floating around the forums to pull the battery for about 15 minutes? Seems to have done the trick for me.
  5. #5  
    I get the same freeze on the black orb with 2 arrows. I'd been getting it on 1.3.5.1 but it went away... some time after removing ez-ipupdate and fudging around with lighttpd to get webdav working. Not a huge help I know.

    I'm now on a doctored 1.4 with lighttpd and webdav but ez-ipupdate has never been on the Pre since doctoring. I can say with certainty that I've had lighttpd installed and running and been able to restart in the past though and both then and now I've been using the same means to set it all up. About the only possible difference I can think of is the version of lighttpd installed. I've got 1.4.25-5 from Optware running at the moment. I may (my memory's not that great!) have had a different version installed before when restarting was working. It would have been pulled down and installed using:
    /opt/bin/ipkg-opt install lighttpd
    But I'm to much of a noob to know how to work out which version that would've installed and exactly where it would've come from... if, in fact, it was any different at all. Helpful huh?!
  6. ferchope's Avatar
    Posts
    92 Posts
    Global Posts
    105 Global Posts
    #6  
    lighttpd puts a LOT of load on the Palm Pre, only when it starts up but doesn't affect performance of the Pre thereafter
  7. #7  
    Quote Originally Posted by sticktime View Post
    Had a similar problem freezing on restart. Did you try the suggestion floating around the forums to pull the battery for about 15 minutes? Seems to have done the trick for me.
    certainly when i pull the battery and put it back in, my pre is able to startup and work just fine. i have not needed to do this for 15 minutes, just a second or two does the trick. what i'm hoping to do is figure out what is hanging the pre on shutdown/reboot so that it can complete the operation gracefully - both for convenience and safety of the fs.
  8. #8  
    Just installed lighty webserver today and have already ssh installed according to palmpre-hacks.com. Did a restart after changing the lighty config file to get rid of the "www" and it restarted just fine, so canīt say I have the same problem. What I donīt have is "ez-ip". Running 1.4 on German GSM Pre.
  9. Tmillz's Avatar
    Posts
    9 Posts
    Global Posts
    14 Global Posts
       #9  
    @ jackx. Thanks 4 looking in to this. More info on my pre. I run lighttpd over evdo on port 8081. And also have the package that alows .php. When I get the chance I will remove the script that updates my ezip when my evdo ip changes and try a restart then.
  10. #10  
    See http://forums.precentral.net/2275484-post5.html for the solution to SSH problems.

    -- Rod
    WebOS Internals and Preware Founder and Developer
    You may wish to donate by Paypal to donations @ webos-internals.org if you find our work useful.
    All donations go back into development.
    www.webos-internals.org twitter.com/webosinternals facebook.com/webosinternals
  11. #11  
    Quote Originally Posted by jackx View Post
    certainly when i pull the battery and put it back in, my pre is able to startup and work just fine. i have not needed to do this for 15 minutes, just a second or two does the trick. what i'm hoping to do is figure out what is hanging the pre on shutdown/reboot so that it can complete the operation gracefully - both for convenience and safety of the fs.
    Pulling the battery did allow me to restart once after startup so I thought the problem was solved. Subsequent restarts hung again however.

    For me the issue was ez-ipupdate. Once I removed that (via preware) I've been able to consistently restart. For now I've switched to updating DDNS with the url service as documented here:

    webos-internals.org/wiki/Tutorials_Linux_DDNS_Update_Client_wget
  12. #12  
    The issue isn't specific to 1.4.0. It existed in 1.3.5.X as well. It isn't in startup. ez-ipupdate is doing something that causes reboot/shutdown to hang. ez-ipupdate is not running in daemon mode, so there must be a odd side effect on the ppp0 device.
  13. #13  
    Quote Originally Posted by dgdonovan View Post
    The issue isn't specific to 1.4.0. It existed in 1.3.5.X as well. It isn't in startup. ez-ipupdate is doing something that causes reboot/shutdown to hang. ez-ipupdate is not running in daemon mode, so there must be a odd side effect on the ppp0 device.
    interesting, i was previously running 1.3.5.1 with a manually install of ez-ipupdate (done using instructions from webos-internals.org wiki) and never had reboot/shutdown problems. on updating to 1.4.0 i also replaced it with the preware package simply to get all my customizations a step closer to being managed by preware. perhaps the issue is a result of the way the preware package is installed?
  14. #14  
    bad news - i removed the ez-ipupdate optware package from my pre and it still hangs on shutdown/reboot, requiring a battery pull to reset. i verified via shell that all the files were removed from the appropriate place. this puts me back to square one on this new problem...
  15. #15  
    I just doctored a Sprint Pre, installed Preware, Lighty and ez-ipupdate.

    Lighty access worked fine.

    It rebooted fine.

    ez-ipupdate does nothing by default until you create an /opt/etc/ez-ipupdate.conf files. I didn't test that any further, as I don't have CDMA in Australia, so cannot get a ppp0 interface to come up.

    So I cannot reproduce these problems, and as far as I can see the packaging is correct.

    -- Rod
    WebOS Internals and Preware Founder and Developer
    You may wish to donate by Paypal to donations @ webos-internals.org if you find our work useful.
    All donations go back into development.
    www.webos-internals.org twitter.com/webosinternals facebook.com/webosinternals
  16. #16  
    I don't believe I've used either (used only Preware with light patches), but I did an EPR before updating to 1.4. A day after updating, my phone got extremely sluggish while texting a using the web. Then I started to get the Too Many Cards error. I exited all apps and was still getting the error every couple of seconds.

    So I hit Orange+Sym+R and 9 mins latter it restarted.
    At the palm (transitioning) it froze and got super hot. Froze at the same place everytime.

    I Doctored it but I still get the "Too Many Cards" error more frequantly than normal, especially while using the Web.

    1.4 seems to make my Pre a lot more sluggish...anyone else experiencing this, even after Doctoring?

    Maybe the Doctoring process went wrong...
  17. RickNY's Avatar
    Posts
    254 Posts
    Global Posts
    319 Global Posts
    #17  
    Quote Originally Posted by rwhitby View Post
    I just doctored a Sprint Pre, installed Preware, Lighty and ez-ipupdate.

    Lighty access worked fine.

    It rebooted fine.

    ez-ipupdate does nothing by default until you create an /opt/etc/ez-ipupdate.conf files. I didn't test that any further, as I don't have CDMA in Australia, so cannot get a ppp0 interface to come up.

    So I cannot reproduce these problems, and as far as I can see the packaging is correct.

    -- Rod
    Hey Rod.. I just did the same as you -- I created the conf file and verified that it did update the DNS properly. I then rebooted the phone once, and it rebooted normally.. I then rebooted a second time, and got the hang at the spinning orb (it stopped spinning and just froze there). Ended up doing a battery pull to reboot.. And yes -- the phone got very hot at the freeze.
  18. #18  
    Quote Originally Posted by RickNY View Post
    Hey Rod.. I just did the same as you -- I created the conf file and verified that it did update the DNS properly. I then rebooted the phone once, and it rebooted normally.. I then rebooted a second time, and got the hang at the spinning orb (it stopped spinning and just froze there). Ended up doing a battery pull to reboot.. And yes -- the phone got very hot at the freeze.
    so we have confirmation that a clean pre with optware, lighty, and ez-ipupdate can lead to this freeze condition. has anyone tried this only installing one of lighty or ez-ipupdate so that we can determine if this is caused by one of these two packages in isolation, or if it is potentially an interaction effect of the two?
  19. plee3ac's Avatar
    Posts
    109 Posts
    Global Posts
    115 Global Posts
    #19  
    I have installed ez-ipupdate using Preware, but never installed the lighty package, and I also get the reset problem (icon stops spinning). I was able to reset the phone using the hold power switch while toggling the ringer switch 3 times.

    When I renamed the "ez-ipupdate.conf" file to "ez-ipupdate.save" to disable the ez-ipupdate process, I was able to reset normally without hanging.

    I guess when I want to enable ez-ipupdate, I will have to rename the conf file back and reboot.

    Hope this helps... plee3
  20. qqq
    qqq is offline
    qqq's Avatar
    Posts
    35 Posts
    #20  
    I'm not sure why the ppp/ifp-up.d/whatever script with ez-ipupdate is causing hangs, but here's what I use in mine, instead of ez-ipupdate:

    Code:
    #! /bin/sh
    date >> /var/log/ipupdate.log
    (sleep 10; /usr/bin/curl --interface ppp0 --retry 6 "http://username:password@members.dyndns.org/nic/update?hostname=myhostname.ath.cx&wildcard=NOCHG&mx=NOCHG&backmx=NOCHG" >> /var/log/ipupdate.log ) &
    It seems to work ok, but I also have no idea why this doesn't break reboots or if that will be the case for anyone else.
Page 1 of 2 12 LastLast

Posting Permissions