Retrspect 5.0 crashes

[icgeb]

Hello,

 

we've reported the following Retrospect failure:

 

Description of the problem:

 

Sometimes Retrospect crashes when the "Devices" button is activated ("Retrospect Directory"

window, and then "Configure" tab selection)

 

 

The system configuration is:

 

Software

 

Retrospect 5.0.238

Mac OS X 10.2.6 (6L60)

 

Hardware

 

PowerMac G4 ( version = 3.3 )

backup device :SCSI DAT Seagate DDS4 mod. 9SP40-000

 

Any suggestion is welcome!!!

 

 

 

The Retrospect crash log is:

 

--------------------------------------------------

 

**********

 

Date/Time: 2003-09-02 13:06:42 +0200

OS Version: 10.2.6 (Build 6L60)

Host: local.test.server

 

Command: Retrospect

PID: 1555

 

Exception: EXC_BAD_ACCESS (0x0001)

Codes: KERN_INVALID_ADDRESS (0x0001) at 0xc0000042

 

Thread 0 Crashed:

#0 0x02130034 in 0x2130034

#1 0x0212d9b4 in 0x212d9b4

#2 0x02132bf0 in 0x2132bf0

#3 0x001d33d4 in 0x1d33d4

#4 0x002522a0 in 0x2522a0

#5 0x001d33d4 in 0x1d33d4

#6 0x002540f0 in 0x2540f0

#7 0x00255064 in 0x255064

#8 0x001d33d4 in 0x1d33d4

#9 0x001a92b8 in 0x1a92b8

#10 0x00225184 in 0x225184

#11 0x00225470 in 0x225470

#12 0x00255314 in 0x255314

#13 0x001d33d4 in 0x1d33d4

#14 0x0027adf4 in 0x27adf4

#15 0x001a8ea0 in 0x1a8ea0

#16 0x001d33d4 in 0x1d33d4

#17 0x001b320c in 0x1b320c

#18 0x001b3250 in 0x1b3250

#19 0x001d33d4 in 0x1d33d4

#20 0x001d5ec4 in 0x1d5ec4

#21 0x001d33d4 in 0x1d33d4

#22 0x00224d5c in 0x224d5c

#23 0x001d33d4 in 0x1d33d4

#24 0x00225cf8 in 0x225cf8

#25 0x001c14a0 in 0x1c14a0

#26 0x001f8e04 in 0x1f8e04

#27 0x001f9000 in 0x1f9000

#28 0x001c19b8 in 0x1c19b8

#29 0x001ccb74 in 0x1ccb74

#30 0x001f8e04 in 0x1f8e04

#31 0x001f9000 in 0x1f9000

#32 0x001cccc0 in 0x1cccc0

#33 0x001f8e04 in 0x1f8e04

#34 0x001f9000 in 0x1f9000

#35 0x001ccdb8 in 0x1ccdb8

#36 0x001ccf04 in 0x1ccf04

 

Thread 1:

#0 0x9000508c in syscall

#1 0x90515d0c in BSD_waitevent

#2 0x905156dc in CarbonSelectThreadFunc

#3 0x90020c88 in _pthread_body

 

Thread 2:

#0 0x9003ea08 in semaphore_wait_signal_trap

#1 0x9003e824 in _pthread_cond_wait

#2 0x9051dbf0 in CarbonOperationThreadFunc

#3 0x90020c88 in _pthread_body

 

Thread 3:

#0 0x90073c08 in mach_msg_trap

#1 0x90005ed0 in mach_msg

#2 0x901489f0 in __CFRunLoopRun

#3 0x90180f58 in CFRunLoopRunSpecific

#4 0x94d9c1c0 in _ZN10HALRunLoop9OwnThreadEPv

#5 0x94d911b0 in _ZN9CAPThread5EntryEPS_

#6 0x90020c88 in _pthread_body

 

Thread 4:

#0 0x9003ea08 in semaphore_wait_signal_trap

#1 0x9003e824 in _pthread_cond_wait

#2 0x90232754 in TSWaitOnSemaphoreCommon

#3 0x90233094 in DeferredTaskThread

#4 0x90020c88 in _pthread_body

 

Thread 5:

#0 0x90073c08 in mach_msg_trap

#1 0x90005ed0 in mach_msg

#2 0x901489f0 in __CFRunLoopRun

#3 0x90180f58 in CFRunLoopRunSpecific

#4 0x0059488c in thTop (thread.c:100)

#5 0x90020c88 in _pthread_body

 

PPC Thread State:

srr0: 0x02130034 srr1: 0x0200f030 vrsave: 0x00000000

xer: 0x20000000 lr: 0x021300d8 ctr: 0x001d95d0 mq: 0x00000000

r0: 0x00000060 r1: 0xbfffde70 r2: 0x01dbba48 r3: 0x02a1bf20

r4: 0x00000030 r5: 0x00499cec r6: 0x06892000 r7: 0x00000000

r8: 0x00046010 r9: 0xa0220a00 r10: 0x90223438 r11: 0x029f8000

r12: 0x0048fc40 r13: 0x00000000 r14: 0x00000000 r15: 0x00000000

r16: 0x00000000 r17: 0xbfffdeb8 r18: 0x00000001 r19: 0x00000001

r20: 0xc0002cb0 r21: 0xffffffff r22: 0x00000001 r23: 0x00000001

r24: 0x0217f79c r25: 0x00000088 r26: 0x01dc33f8 r27: 0x02a1bf50

r28: 0x01dc33f8 r29: 0xbfffded8 r30: 0xc0000040 r31: 0x01dc3400

 

 

[CallMeDave]

Quote:

---

Hardware

 

PowerMac G4 ( version = 3.3 )

backup device :SCSI DAT Seagate DDS4 mod. 9SP40-000

 

 

---

 

Since the PowerMac G4 doesn't have built-in SCSI support, you must be using some sort of host adapter.

 

- What SCSI card are you using?

 

 

Dave

[icgeb]

The SCSI adapter is an "Adaptec 29160"

 

ROM#1.2 rev.2

 

 

 

 

 

Diego

[natew]

Hi

 

How many pins on the Backup device? Are you using the 1.2 version of Adaptecs driver?

 

Nate

[icgeb]

Hi,

 

 

 

version is 1.2 (revision 2)

 

 

 

What do you mean for "pins"? If you intend SCSI connector, the DAT tape unit is external connected

 

by a High Density 68pin SCSI cable.

 

 

 

 

 

Diego

[natew]

Hi

 

 

 

There are often problems when using 68 pin adapters with 50 and 25 pin SCSI devices. There are so many different ways to refer to SCSI connections pin count is one way to make sure we are talking about the same thing.

 

 

 

I assume this works perfectly in OS9? You might want to check to make sure. If at all possible I would try a non-Adaptec SCSI card. Not that the card itself is bad, there have just been tons of problems with Adaptec support in OSX.

 

 

 

Nate

[icgeb]

Hi Nate,

 

The tape unit has a 68 pin wide connector identical to the Adaptec scsi controller board connector;

so we used a "68 to 68" pins wide cable; no cable adapters.

 

It is not possible to test with OS 9 because the PowerMac G4 "Mirrored doors"( version = 3.3 )

boots Mac OS X only.

 

We purchased an Adaptec 29160 board because is a supported adapter as reported by Dantz Article

ID:27381

( http://www.dantz.com/index.php3?SCREEN=kbase&ACTION=KBASE&id=27381 )

 

Diego.

[natew]

Hi

 

I know its on the list but I have seen more problems with that card than any other. Maybe it is just because they are more popular than other cards? Hard to say.

 

At this point the only thing left is to make sure the hardware is OK. Is there any way you can hook this up to another machine?

 

Nate

[sekolec]

I have the same problem, but Retrospect can crash any time - also in the middle of a backup. A recent upgrade to 5.1 made no difference.

 

System: Mac OS Server 10.2.6 (6L60)

Hardware: Xserve dual processor

Backup to: Soft Raid (3 x 172.5 GB)

 

The machine is used also for file sharing.

Is it possible to "auto restart" Retrospect?

 

Luka

[icgeb]

Nate,

 

 

 

Unfortunatelly not, we can halt the system and we do not have any other PowerMac available to test

 

the SCSI adapter.

 

 

 

The system runs as server, and several backup jobs are scheduled to run daily.

 

During backup we Retrsopect never crashed.

 

 

 

 

 

 

 

 

 

Diego

[icgeb]

We got another Retrospect crash, but this time while retrospect was running a scheduled backup.

 

THE BACKUP JOB WAS USING A "FILE BACKUP SET" AND NOT TAPE BACKUP SET.

 

This means that SCSI controller and Tape Unit are not involved in the crash.

 

I'll append the crash report ( it looks very similar to the previus one you find in this thread).

 

**********

 

Date/Time: 2003-09-11 13:50:58 +0200

OS Version: 10.2.6 (Build 6L60)

Host: local.test.server

 

Command: Retrospect

PID: 2958

 

Exception: EXC_BAD_ACCESS (0x0001)

Codes: KERN_INVALID_ADDRESS (0x0001) at 0xc000000a

 

Thread 0 Crashed:

#0 0x0212f034 in 0x212f034

#1 0x0212c9b4 in 0x212c9b4

#2 0x02131bf0 in 0x2131bf0

#3 0x02132078 in 0x2132078

#4 0x001d33d4 in 0x1d33d4

#5 0x00252f00 in 0x252f00

#6 0x0022bec8 in 0x22bec8

#7 0x0022c174 in 0x22c174

#8 0x001d33d4 in 0x1d33d4

#9 0x0026beb4 in 0x26beb4

#10 0x0027029c in 0x27029c

#11 0x001d33d4 in 0x1d33d4

#12 0x0029ab98 in 0x29ab98

#13 0x001d33d4 in 0x1d33d4

#14 0x002aeea8 in 0x2aeea8

#15 0x001d33d4 in 0x1d33d4

#16 0x001f8eac in 0x1f8eac

#17 0x001f9000 in 0x1f9000

#18 0x001ccdb8 in 0x1ccdb8

#19 0x001ccf04 in 0x1ccf04

 

Thread 1:

#0 0x9000508c in syscall

#1 0x90515d0c in BSD_waitevent

#2 0x905156dc in CarbonSelectThreadFunc

#3 0x90020c88 in _pthread_body

 

Thread 2:

#0 0x9003ea08 in semaphore_wait_signal_trap

#1 0x9003e824 in _pthread_cond_wait

#2 0x9051dbf0 in CarbonOperationThreadFunc

#3 0x90020c88 in _pthread_body

 

Thread 3:

#0 0x90073c08 in mach_msg_trap

#1 0x90005ed0 in mach_msg

#2 0x901489f0 in __CFRunLoopRun

#3 0x90180f58 in CFRunLoopRunSpecific

#4 0x94d9c1c0 in _ZN10HALRunLoop9OwnThreadEPv

#5 0x94d911b0 in _ZN9CAPThread5EntryEPS_

#6 0x90020c88 in _pthread_body

 

Thread 4:

#0 0x9003ea08 in semaphore_wait_signal_trap

#1 0x9003e824 in _pthread_cond_wait

#2 0x90232754 in TSWaitOnSemaphoreCommon

#3 0x90233094 in DeferredTaskThread

#4 0x90020c88 in _pthread_body

 

Thread 5:

#0 0x90073c08 in mach_msg_trap

#1 0x90005ed0 in mach_msg

#2 0x901489f0 in __CFRunLoopRun

#3 0x90180f58 in CFRunLoopRunSpecific

#4 0x0059488c in thTop (thread.c:100)

#5 0x90020c88 in _pthread_body

 

PPC Thread State:

srr0: 0x0212f034 srr1: 0x0200f030 vrsave: 0x00000000

xer: 0x20000000 lr: 0x0212efe8 ctr: 0x001d95d0 mq: 0x00000000

r0: 0x00000018 r1: 0xbfffde80 r2: 0x01dbba48 r3: 0x00007830

r4: 0x00001800 r5: 0x00499cec r6: 0x00000008 r7: 0xbfffdec8

r8: 0x00046010 r9: 0xa0220a00 r10: 0x90223438 r11: 0x0228e000

r12: 0x0048fc40 r13: 0x00000000 r14: 0x00000000 r15: 0x00000000

r16: 0x00000000 r17: 0xbfffdec8 r18: 0x00000000 r19: 0x00000000

r20: 0xc0004160 r21: 0x04e58989 r22: 0x00000001 r23: 0xbfffeab0

r24: 0x020c2ec4 r25: 0x00000466 r26: 0x01dc33f8 r27: 0x02295020

r28: 0x01dc33f8 r29: 0xbffff178 r30: 0xc0000008 r31: 0x01dc3400

 

 

 

[sekolec]

I get a similar crash report, below is the first part of it:

 

Date/Time: 2003-09-12 09:20:41 +0200

OS Version: 10.2.6 (Build 6L60)

Host: log000a

 

Command: Retrospect

PID: 13344

 

Exception: EXC_BAD_ACCESS (0x0001)

Codes: KERN_INVALID_ADDRESS (0x0001) at 0x43c87e74

 

Thread 0 Crashed:

#0 0x001fd0ec in 0x1fd0ec

#1 0x001fd440 in 0x1fd440

#2 0x001fd674 in 0x1fd674

 

Retrospect crashes again and again, any time during backup of various clients. At the moment I have 21 Mac OS X clients.

 

I hope somebody has some idea as to what is wrong!

 

Luka

 

[hume]

I too started to get this tonight. After a little repeating I found it scanning my hard disk with approx 70,000 folders and 362,000 files just before the crash. Deleting stuf from the Hard drive has allowed the backup server to work! I hope Dantz has tested large volumes??

 

HTH,

Steve Hume

 

My partial log.

 

Exception: EXC_BAD_ACCESS (0x0001)

Codes: KERN_INVALID_ADDRESS (0x0001) at 0xc44e7830

 

Thread 0 Crashed:

#0 0x002133dc in 0x2133dc

#1 0x003836c0 in 0x3836c0

#2 0x00238e04 in 0x238e04

#3 0x00239000 in 0x239000

 

[sekolec]

I have several users with lots of files, and Retrospect does not always crash with the same user. After restarting Retrospect, the user that "caused" the crash is backed up quite happily.

 

I have found other similar crash reports; perhaps it is an Apple problem. Maybe somebody who understands crash reports could have a look. Below are crash reports from FileServer and ServerManager:

 

Date/Time: 2003-09-12 13:53:11 +0200

OS Version: 10.2.6 (Build 6L60)

Host: log000a

 

Command: AppleFileServer

PID: 463

 

Exception: EXC_BAD_ACCESS (0x0001)

Codes: KERN_INVALID_ADDRESS (0x0001) at 0x465053ed

 

Backtrace not available

 

PPC Thread State:

srr0: 0x00019db4 srr1: 0x0000f030 vrsave: 0x00000000

xer: 0x20000000 lr: 0x0004c04c ctr: 0x000506e4 mq: 0x00000000

r0: 0x9015bd70 r1: 0xf0080600 r2: 0x00000024 r3: 0xf00807e0

r4: 0x465053e9 r5: 0x00000028 r6: 0x001ba960 r7: 0x00000000

r8: 0x465053e9 r9: 0xa0001048 r10: 0xa0130d68 r11: 0xa0132e30

r12: 0x000506e4 r13: 0x00000000 r14: 0x00000000 r15: 0x00000000

r16: 0x00000001 r17: 0x00000000 r18: 0x24002280 r19: 0x00000000

r20: 0x00001f07 r21: 0x00000000 r22: 0x001bb798 r23: 0xf0080c34

r24: 0x00000028 r25: 0xf00807e0 r26: 0xa012bc90 r27: 0x001ba960

r28: 0x46505365 r29: 0xf00807e0 r30: 0x00000000 r31: 0x465053e9

 

 

Date/Time: 2003-09-15 23:07:02 +0200

OS Version: 10.2.6 (Build 6L60)

Host: log000a

 

Command: servermgr_info

PID: 8672

 

Exception: EXC_BAD_ACCESS (0x0001)

Codes: KERN_INVALID_ADDRESS (0x0001) at 0xa1b1c1d3

 

Thread 0 Crashed:

#0 0x9068ba4c in objc_msgSend

#1 0x97e1ac38 in _replacementObjectForObject

#2 0x97e14a78 in _encodeObject

#3 0x97e2894c in -[NSArray encodeWithCoder

#4 0x97e104c8 in -[NSDictionary encodeWithCoder

#5 0x97e4acec in -[NSArchiver encodeRootObject

#6 0x97e4ac00 in +[NSArchiver archivedDataWithRootObject

#7 0x00004a30 in 0x4a30

#8 0x00005518 in 0x5518

#9 0x000026ec in 0x26ec

#10 0x0000256c in 0x256c

 

PPC Thread State:

srr0: 0x9068ba4c srr1: 0x0000f030 vrsave: 0x00000000

xer: 0x00000000 lr: 0x97e1ac38 ctr: 0x9068ba3c mq: 0x00000000

r0: 0x97e1ac38 r1: 0xbffff050 r2: 0x28000241 r3: 0xa1b1c1d3

r4: 0x906cbba4 r5: 0x0005d540 r6: 0x000041d3 r7: 0xa1b1c1d3

r8: 0x0000457a r9: 0x000003a7 r10: 0xa1b1c1d3 r11: 0xa7df1638

r12: 0x9068ba3c r13: 0x00000000 r14: 0x00000000 r15: 0x00000000

r16: 0x00000000 r17: 0x00000000 r18: 0x00000000 r19: 0x00000000

r20: 0x00000000 r21: 0x00000000 r22: 0x00000000 r23: 0x00000000

r24: 0x00000000 r25: 0x0005f9d0 r26: 0x0005d540 r27: 0x0005f9d0

r28: 0x00175ae0 r29: 0xa1b1c1d3 r30: 0x0005d540 r31: 0x97e1abd4

 

 

[sekolec]

Retrospect seems to run happily for a couple of days after rebooting the server; indicates that Mac OS X/Retrospect is running out of resources.

 

Can somebody provide a script to reboot the server at regular intervals? Not a very clever solution, but maybe worthwhile, if it solves the problem.

 

Luka

[natew]

Hi

 

Is Retrospect running on the server you want to reboot? If so you can set Retrospect to reboot after scripts complete.

 

Nate

[Mr_Martini]

I have a similar problem

 

Mac Xserve, OSX Server 10.2.6, Sony 4 cartridge AIT drive, Retropsect 5.1.175 Workgroup

 

Problem - Retrospect quits out during auto-launch scripts, nearly always on the second disk of the backup session. It does not matter in which order the disks are backed up

 

Running the script manually works at least 95% of the time

 

I have also seen this same problem on Mac G4's running Retropsect 5 in OS9, except it totally freezes theses computers. Retropsect 4 running on the same Macs has no problems

 

Any ideas anyone??

 

Mr Martini

[sekolec]

Yes, Retrospect is running on the server I want to reboot.

 

I have set the option to restart after completion, but it is ignored probably due to the fact that the script for client backup is still hanging around.

 

I have two scripts: one for backing up the the network clients and one for backing up the server itself. The server gets backed up during the night. The client backup is always active, except during the time reserved for the server backup. I presume that, although the client backup is inactive, it is still effectively running in the background, so the restart does not take place.

 

A restart would be a partial solution, better would be to find out the cause of the crashes. Any ideas from Dantz?

 

Luka

[sekolec]

My configuration:

 

Xserve G4 dual processor

4 x 256 MB RAM

4 x 180GB disks - 1 for System and file services

- 3 for backup (Apple Soft Raid)

Retrospect 6

 

After months of struggling with Retrospect (version 5 and 6) crashes and elem.c errors, I started manipulating the memory configuration. I found that one 256 MB RAM was not quite compatible with two others. Since removing the culprit, Retrospect runs for weeks on end. This memory problem was evident only with Retrospect, no other programme.