Unexpected log-outs!

rinse

rinse

www.visualrinse.com
WTF!!!!!

I'm working, browsing the web a bit.... general computing.

Then, BAMM! the screen goes pale blue and im thrown into my loging screen. Without choosing to log out.

I have OS 10.1.2, IE 5.1 (Omniweb didn't fix the problem, i thought maybe IE was playing crazy games on me), and i recently installed MySQl... I dont know if that would have done anything

Anyone experinecing similar things?

The Console says nothing of the event... how can I find out what is happening? *naive* Am I getting hacked?

this is most distressing.

i am turning off all sharing stuff. maybe that will fix things up.

:confused:
 
I was having the same exact problem and finally I figured out that the 512MB DIMM I got from Outpost.com was bad. I replaced it with a DIMM from Crucial.com and the problem went away.
 
killing my windowserver PID does the exact same effect... crap.

i also have a 512 mb module from Outpost... i hope that isn't it. i like having that much ram and can't afford more right now.

are there any freeware apps to run memory checks ? something like techtool pro's memory test but free?

Tommy, have you contacted Outpost about their memory's problems?
 
Originally posted by testuser
Call up Outpost and describe to them the problems you are having. Tell them that you suspect it is related to their memory module. (You could verify this by removing the module, and running your computer for a while without it. If it does not crash without the memory, then you have a good indication that the memory is bad.) Most vendors have warranties on their RAM, and will be nice enough to ship you a replacement module in return for your defective one.

By the way, software RAM checkers will not necessarily catch all faulty RAM. I would not rely on a software diagnosis. The only way to reliably tell if the RAM is bad is to see how the machine functions with it, and without it.
Click to expand...
Yeah... I did that.

I did not mean to imply that Outpost.com gave me a hard time. They didn't. They gave me a refund with no problem.

BTW, with my new G4 I got the 3 year Apple Care. It came with a hardware check CD. That's what identified this as a bad DIMM...
 
i ran fsck -y .....i had some invalid keys and fixed them.

and then i ran the hardware test cd (which came with my non-apple care G4) and the three memory chips (including the 512 from outpost) all passed.

i haven't experienced the Window Server Crash fro about an hour (whew!)

ill post again at the end of the day to let you guys know if it is stilll messed up.
 
For future reference, there is a very useful freeware RAM checker available from Versiontracker (and probably several other sites). It's called DIMMCheck, and it's a Classic utility. I used it before running that infamous Destroy-3rd-party-RAM firmware update, and it correctly told me that I had nothing to worry about.
:D
 
it is still doing it! gah!

i wish there was a way to log Window Manager Crashes.... Console doesn't do it...

im gonna remove that chip and see how it performs over the next day or so.

this is annoying as hell.
 
You can type 'last' in the terminal to see the last few logins to your computer.. if something shows up from an IP you are not familiar with, I'd start to worry.

netstat -f inet
this will show you all your network connections, so try running this directly after you are logged out.

Obviously, disconnecting your network cables will eliminate any external connections better than anything, so you could try that.

Does this happen regardless of the websites you are visiting?

Here's to hoping you get this figured out.
 
Oh yeah, the log files that 'last' reads from are regularly circulated so they don't get outrageously long. To see your old logs:
[mach4:~] kilowatt% cd /var/log
[mach4:/var/log] kilowatt% ls -l | grep wtmp
-rw-r----- 1 root wheel 468 Jan 3 17:23 wtmp
-rw-r----- 1 root wheel 1168 Dec 31 19:21 wtmp.0.gz
-rw-r----- 1 root wheel 2002 Nov 30 15:32 wtmp.1.gz
-rw-r----- 1 root wheel 1342 Nov 1 01:29 wtmp.2.gz
-rw-r----- 1 root wheel 606 Sep 30 19:22 wtmp.3.gz
-rw-r----- 1 root wheel 310 Aug 31 2000 wtmp.4.gz
[mach4:/var/log] kilowatt%
[mach4:/var/log] kilowatt% cp wtmp.1.gz ~/wtmp.1.gz ; gunzip ~/wtmp.1.gz ; last -f ~/wtmp.1 ; rm ~/wtmp.1
Click to expand...

This should print out tons of logins. Replace wtmp.1.gz with any other wtmp files you want to see.
 
Originally posted by rinse
WTF!!!!!

I'm working, browsing the web a bit.... general computing.

Then, BAMM! the screen goes pale blue and im thrown into my loging screen. Without choosing to log out.

I have OS 10.1.2, IE 5.1 (Omniweb didn't fix the problem, i thought maybe IE was playing crazy games on me), and i recently installed MySQl... I dont know if that would have done anything

Anyone experinecing similar things?

The Console says nothing of the event... how can I find out what is happening? *naive* Am I getting hacked?

this is most distressing.

i am turning off all sharing stuff. maybe that will fix things up.

:confused:
Click to expand...

I had this exact same symptom but my issue was a faulty daughter-card in my Pismo. I found an unused Pismo and swapped daughter cards and resolved the issue. This symptom seems to be the result of any number of issues and my three weeks of hell with Apple support were fruitless (As they'd return it each time saying it was fine, hence my scavaging an old Pismo)
 
i never ended up removing the memory, but the problem hasn't occured for the last day or so.... knock on wood.
 
ok, so it kept happening...

then i got the idea...

well if the windowserver is crashing.... maybe if i trash the windowserver's prefs it will help!

so i deleted library/preferences/com.apple.windowserver.plist and things seem to be running a bit smoother.

we'll see over the next couple days.
 
well... after 1 day and 10 hours of uptime, i have not experienced a windowserver crash after trashing the windowserver's prefs!

yay! :)

as a sidenote, this also fixed some monitor resolution issues i was having on reboots.

double yay! :) :)
 
so... it happened again. and i fianlly rmeove d the 512 chip from outpost. and then it did it again. and now i have removed the 256 chip i bought from Memory X. we will what happens now...

i can't think it would be the 128 chip from apple.... no way.

they only other card / hardware i have plugged into the machine is an extra nic card i use to connect with a Win2k PC....


this is really pissing me off. :mad: :mad: :mad: :mad: :mad: :mad:
 
okay...

users/chad/library/logs/windowserver.crash.log


the text from a crash...

Date/Time: 2002-01-10 18:22:50 -0600
OS Version: 10.1.2 (Build 5P48)

Command: WindowServer
PID: 70

Exception: EXC_BAD_ACCESS (0x0001)
Codes: KERN_PROTECTION_FAILURE (0x0002) at 0x00000090

Thread 0:
#0 0x77182470 in CGSReleaseRegion
#1 0x77176fe8 in CGXRedrawDisplay
#2 0x771990a8 in _CGXFlushWindowContent
#3 0x77198b68 in CGXFlushWindow
#4 0x771a015c in _CGXFlushRegionInline
#5 0x771ae980 in _XFlushRegionInline
#6 0x7719d34c in CGXServices_server
#7 0x77198d38 in connectionHandler
#8 0x7719360c in CGXPostPortData
#9 0x77177d24 in CGXRunOneServerPass
#10 0x7717c4f0 in CGXRunOneServicesPass
#11 0x77180978 in CGXServerLoop
#12 0x7729d8d4 in CGXServer
#13 0x00002ecc in main
#14 0x00002e08 in _start
#15 0x00002c38 in start

Thread 1:
#0 0x70000978 in mach_msg_overwrite_trap
#1 0x70005b34 in mach_msg_overwrite
#2 0x771e4064 in asyncThread
#3 0x7002054c in _pthread_body

Thread 2:
#0 0x7003f4c8 in semaphore_wait_signal_trap
#1 0x772c0414 in layer_blit_thread
#2 0x7002054c in _pthread_body


PPC Thread State:
srr0: 0x77182470 srr1: 0x0200d030 vrsave: 0x00000008
xer: 0x0000000c lr: 0x77182464 ctr: 0x700056fc mq: 0x00000000
r0: 0x77176fe8 r1: 0xbffef650 r2: 0x00000000 r3: 0x00000090
r4: 0x00000000 r5: 0x00000011 r6: 0x00000000 r7: 0x00000803
r8: 0x0043ea20 r9: 0x00000000 r10: 0x0008e1a0 r11: 0x80003710
r12: 0x771934ac r13: 0x00000000 r14: 0x00000036 r15: 0x0005c400
r16: 0x00000001 r17: 0x80160fa8 r18: 0x00060ef8 r19: 0x00001b07
r20: 0x00000000 r21: 0x0000001c r22: 0x70004234 r23: 0x700042c8
r24: 0x7016b0dc r25: 0x006bac3c r26: 0x8081ab5c r27: 0xc0d51800
r28: 0x00000000 r29: 0xbfffef00 r30: 0x00000000 r31: 0x00000001

**********



any ideas?
 
this machine is only two months old, so everything is still stock.

1 40 gig stock hard drive
cd-rw
additional nic card (i think it is intel, but not sure)

the only driver i have installed is for a wacom tablet...


as far daemons and background apps...

samba
MySQL
Apache

after pulling that 256 meg chip things seem to working fine (for the last 18 hours or so) we'll how it goes.

thanks for you input in all of this. i 'll keep everyone posted.
 
Well, it's not a solution to your problem now, but have you submitted this as a bug to apple? maybe including the crash log would help them a bit...
 
I'm starting to have this problem. I'm hoping it's not a 512 chip i bought a while ago, it passed a test by techtool pro, so i'm not sure. I know I get logged out every so often, maybe twice a week, if that. I get that ugly text over the interface crash every other day. I fscked, and the system file was modified. I ran the permission repair, and that found all types of things, it doesn't bomb out as much now, but I get logged out every so often. I think it's time to jump to 10.2
 
Originally posted by Urbansory
I'm starting to have this problem. I'm hoping it's not a 512 chip i bought a while ago, it passed a test by techtool pro, so i'm not sure. I know I get logged out every so often, maybe twice a week, if that. I get that ugly text over the interface crash every other day. I fscked, and the system file was modified. I ran the permission repair, and that found all types of things, it doesn't bomb out as much now, but I get logged out every so often. I think it's time to jump to 10.2
Click to expand...


I had this problem twice. The first time was a bad daughtercard. The card passed all of Apple's diagnostics (they even kept sending the powerbook back each time I sent it) but they for some reasn never actually used the machine to see the error. swapping with an identical model card solved my problem with that.

The second time was b/c of one of my 256MB chips. Removing one chip resolved the problem (it was the chip b/c I replaced the good one with it to be sure it wasn't the two together, and the bad chip had problems even by itself). My bad chip also passed the Techtool diag.

I'd suggest taking the 512 out and just seeing if the problem persists. If it does, and the chip is new, send it back to the manufactuer.
 
I'll give it a try. I tried Dimm Firstaid, and all my dimms passed. I didn't start having real problems till last week. So I reinstalled my entire system, and proir to installing my apps, I had a crash, the one with text over the interface.

I aslso noticed crashes happened when I came from a monitor sleep, or when the screensaver was running. Recently, the screensave flickered, off and on and had a distorted look, like corrupted files. My screensaver images are on another partitin, which i scanned with Norton, and they passed. So I restarted, fscked in single user, and that problem was gone, and again their was a modification to the system somehow, so that was fixed. Just yesterday, right before it logged me out, a few apps kept quiting, so i guess it made it unstable and the system eventually caved and kicked me out.

Is there a way to make sure my ram is good? I had this ram for a month now, and prior to installing it, i had problems, not like this though. I had a crash every now and then, it's not everyday, but I have them. I have to restart to burn CD in 9, so my uptime is always cut short.
 
You must log in or register to reply here.
Back
Top