Jump to content

SCSITaskUserClient errors fill system log until kernel panic


Recommended Posts

I am running Retrospect 6.1.230 with driver update 6.1.16.100 on a 1.25 GHz Mac mini G4 with 1 GB RAM, and an external connected Firewire Exabyte VXA2 tape drive.

 

While Retrospect is performing backups to the tape drive, the following errors are constantly output to the system log:

 

May 9 20:59:05 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 9 20:59:35: --- last message repeated 4 times ---

May 9 21:00:39 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 9 21:01:38 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 9 21:02:08: --- last message repeated 4 times ---

May 9 21:02:29 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 9 21:02:59: --- last message repeated 4 times ---

May 9 21:03:23 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 9 21:03:53: --- last message repeated 4 times ---

May 9 21:04:35 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 9 21:05:05: --- last message repeated 4 times ---

May 9 21:05:28 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 9 21:05:58: --- last message repeated 4 times ---

May 9 21:06:19 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 9 21:06:49: --- last message repeated 4 times ---

May 9 21:07:04 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 9 21:07:34: --- last message repeated 4 times ---

May 9 21:07:46 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 9 21:08:16: --- last message repeated 4 times ---

May 9 21:08:29 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 9 21:08:59: --- last message repeated 4 times ---

May 9 21:09:12 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 9 21:09:42: --- last message repeated 4 times ---

May 9 21:09:58 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 9 21:10:28: --- last message repeated 4 times ---

May 9 21:10:41 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 9 21:11:11: --- last message repeated 4 times ---

 

After days of this, the machine eventually kernel panics. Here's a sample panic log from today:

 

$ cat 2010-05-09-090349.panic

Sun May 9 09:03:49 2010

panic(cpu 0 caller 0x00044040): "zalloc: \"threads\" (135171 elements) retry fail 3"@/SourceCache/xnu/xnu-1228.15.4/osfmk/kern/zalloc.c:966

Latest stack backtrace for cpu 0:

Backtrace:

0x0009CC88 0x0009D63C 0x00029DA0 0x00044040 0x0003D034 0x0003D3F8 0x0003D490 0x00337D38

0x24F01CA0 0x24F02E8C 0x24F02F70 0x24EF8554 0x24F000F8 0x24F0257C 0x24F023EC 0x0003F2A0

0x000B1DD4

Kernel loadable modules in backtrace (with dependencies):

com.apple.iokit.IOATABlockStorage(2.0.6)@0x24ef5000->0x24f07fff

dependency: com.apple.iokit.IOStorageFamily(1.5.6)@0x24e88000

dependency: com.apple.iokit.IOATAFamily(2.0.1)@0x24789000

Proceeding back via exception chain:

Exception state (sv=0x1e689000)

PC=0x00000000; MSR=0x0000D030; DAR=0x00000000; DSISR=0x00000000; LR=0x00000000; R1=0x00000000; XCP=0x00000000 (Unknown)

 

BSD process name corresponding to current thread: kernel_task

 

Mac OS version:

9L31a

 

Kernel version:

Darwin Kernel Version 9.8.0: Wed Jul 15 16:57:01 PDT 2009; root:xnu-1228.15.4~1/RELEASE_PPC

System model name: PowerMac10,1

 

System uptime in nanoseconds: 298675238810105

vm objects:2090900

kernel map entries:1417196

kalloc.32:1515520

kalloc.64:2523136

vm pages:11264924

ipc ports:19232220

threads:126520056

uthreads:56234880

vnodes:2509200

HFS node:3908128

HFS fork:1577968

buf.8192:42729472

Kernel Stacks:1228800

Kalloc.Large:2139441

unloaded kexts:

com.apple.driver.Apple02DBDMAAudio 2.5.8f1 - last unloaded 5974866976

loaded kexts:

com.apple.driver.AppleToonieAudio 2.5.8f1 - last loaded 2369254808

com.apple.filesystems.autofs 2.0.2

com.apple.driver.AppleOnboardAudio 2.5.8f1

com.apple.driver.AppleTexasAudio 2.5.8f1

com.apple.driver.DiskImages 199

com.apple.driver.InternalModemSupport 2.4.0

com.apple.driver.AppleTexas2Audio 2.5.8f1

com.apple.driver.MotorolaSM56K 1.3.9

com.apple.driver.AppleSCCSerial 1.3.2

com.apple.driver.AppleDACAAudio 2.5.8f1

com.apple.driver.AudioIPCDriver 1.0.6

com.apple.iokit.AppleGMACEthernet 1.5.9f1

com.apple.iokit.IOFireWireIP 1.7.7

com.apple.ATIRadeon8500 5.4.8

com.apple.driver.AppleThermal 1.0.1f2

com.apple.driver.AppleSEP 1.3.0

com.apple.driver.AppleK2Driver 1.7.2f1

com.apple.driver.AppleHWClock 1.5.2d0

com.apple.driver.AppleI2S 1.0.1f1

com.apple.iokit.IOSCSIMultimediaCommandsDevice 2.1.1

com.apple.iokit.SCSITaskUserClient 2.1.1

com.apple.driver.AppleFileSystemDriver 1.1.0

com.apple.driver.LSI_FW_500 2.0.9

com.apple.driver.Oxford_Semi 2.0.9

com.apple.driver.StorageLynx 2.0.9

com.apple.driver.initioFWBridge 2.0.9

com.apple.driver.IOFireWireSerialBusProtocolSansPhysicalUnit 2.0.9

com.apple.driver.XsanFilter 2.7.91

com.apple.iokit.IOATAPIProtocolTransport 1.5.3

com.apple.iokit.IOATABlockStorage 2.0.6

com.apple.driver.AppleUSBHub 3.4.9

com.apple.iokit.IOUSBUserClient 3.5.2

com.apple.driver.ApplePMU 2.5.6d2

com.apple.driver.AppleFWOHCI 3.9.7

com.apple.driver.AppleMPIC 1.5.3

com.apple.driver.AppleGPIO 1.3.0d0

com.apple.driver.AppleKauaiATA 1.2.1f4

com.apple.driver.

 

As you can see, the panic happens with com.apple.iokit.IOATABlockStorage as the most recent module in the backtrace.

 

I've done a little research on this error, and according to what I've read on Apple's SCSI mail list, it is due to improperly written SCSI software:

 

http://discussions.apple.com/thread.jspa?threadID=2006207&tstart=0

 

There's nothing special about my setup. It's a PPC Mac mini connected to an external Exabyte Firewire tape drive, running the latest Retrospect 6.1.

 

Please advise!

Link to comment
Share on other sites

NOTE: First, I forgot to mention that the VXA-2 Firewire tape drive I am using has the latest firmware (vxa2_v12153.zip) installed, and is running fine.

 

Second, I included the wrong link in my post above. This is the link to the Apple ata-scsi-dev mail list archive where they discuss the cause of these messages:

 

http://lists.apple.com/archives/ata-scsi-dev/2009/Aug/msg00007.html

 

In this mail list thread, Apple says the messages are sent to the system log because:

 

"the application is trying to send a SCSI Task and has not changed the Scatter-Gather entries. The task must have previously had a S/G entry and now does not (but the application did not reset the S/G list).

 

In 10.6, there is a new API (ResetForNewTask()) to help you do this. In the meantime, you should be able to call SetScatterGatherEntries() and pass 0 for the number of entries."

 

It seems to me that Retrospect is at fault here, because the messages appear only while Retrospect is accessing the tape drive, and not at any other time.

 

I should also mention that these messages seem to be posted to the system log with much greater frequency while Retrospect is waiting for a new tape. For instance, here's what my system log looks like this morning (just one day after the last kernel panic) while it is waiting for a new tape to be inserted:

 

May 10 08:39:12: --- last message repeated 109 times ---

May 10 08:39:12 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 10 08:39:42: --- last message repeated 107 times ---

May 10 08:39:42 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 10 08:40:12: --- last message repeated 107 times ---

May 10 08:40:12 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 10 08:40:42: --- last message repeated 109 times ---

May 10 08:40:42 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 10 08:41:12: --- last message repeated 109 times ---

May 10 08:41:12 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 10 08:41:42: --- last message repeated 107 times ---

 

Eventually (probably soon), the machine will kernel panic - something I am not looking forward to!

 

Please advise!

Edited by Guest
Link to comment
Share on other sites

So is it unreasonable to expect EMC to read this post and reply to it?

Actually, yes. These are user-to-user support forums, and are not the way to contact EMC support. We are just Retrospect users like you are.

 

See: Retrospect Forum Rules/Terms of Use

 

This forum is a community based self help tool for users of the Retrospect Backup Software and other EMC Insignia Products. ...

 

While this forum is monitored by members of the EMC Technical Support team, it is not possible to reply to all questions and threads. This forum is not an official method for contacting technical support. EMC employees are under no obligation to reply to individual forum posts. If you need immediate technical support, you can contact technical support directly at http://www.emcinsignia.com/contactsupport

I suggest that you contact EMC Retrospect support at the link above.

 

If this is a bug / problem with Retrospect 6.x, don't expect for it to be fixed. That product is dead, end-of-life.

 

All development work is now on Retrospect 8.

 

Hope this helps clear up your misunderstanding.

 

Russ

Link to comment
Share on other sites

Actually, Snow Leopard compatibility is not at issue here, so that link is irrelevant.

 

I am running Retrospect 6.1.230 with driver update 6.1.16.100 on a 1.25 GHz Mac mini G4 with 1 GB RAM
There's nothing special about my setup. It's a PPC Mac mini ... running the latest Retrospect 6.1.

Now, while it would have been nice if the original poster had provided details of exactly what version of Mac OS X was running on the computer exhibiting these errors, we can infer that it cannot be Snow Leopard because Snow Leopard only runs on Intel architecture Macs, and it does not run on a PowerPC computer.

 

Russ

Link to comment
Share on other sites

I'm running 10.5.8 with all software updates applied.

 

And, trust me, I wouldn't be using Retrospect 6 if version 8 wasn't so horribly unstable, problematic, and completely unreliable!!!

 

Grr... Sometimes I am very sorry I ever bought Retrospect 8. If I hadn't purchased this tape drive, I would have switched to another backup software package long ago!

Edited by Guest
Link to comment
Share on other sites

And, trust me, I wouldn't be using Retrospect 6 if version 8 wasn't so horribly unstable, problematic, and completely unreliable!!!

You and I are in complete agreement there, except you left out "feature lacking" (can't read older Retrospect backups; can't be scripted, which makes it impossible to coordinate server service shutdown/checkpoint/restart with Retrospect backups).

 

But I am hopeful that the "big bugfix update" expected later this quarter century will improve things quite a bit.

 

Russ

Link to comment
Share on other sites

Perhaps. However, it would seem to be specific to the Firewire version of your drive. I've got the SCSI version of the same drive, with autoloader (attached to an ATTO UL4D) and have never seen this error with Retrospect 6.1.230, RDU 6.1.16.100, on our Xserve.

 

russ

Link to comment
Share on other sites

Looks like I spoke too soon. Once Retrospect started another backup this morning, the messages are back:

 

May 21 09:44:34 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 21 09:45:04: --- last message repeated 2 times ---

May 21 09:46:38 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 21 09:47:34 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 21 09:48:04: --- last message repeated 4 times ---

May 21 09:48:30 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 21 09:49:00: --- last message repeated 4 times ---

May 21 09:49:26 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 21 09:49:56: --- last message repeated 4 times ---

May 21 09:50:20 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 21 09:50:50: --- last message repeated 4 times ---

May 21 09:51:28 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 21 09:51:58: --- last message repeated 4 times ---

May 21 09:52:23 backup kernel[0]: SCSITaskUserClient - Invalid arguments: scatterGatherEntries = 1, requestedTransferCount = 0, transferDirection is 0

May 21 09:52:53: --- last message repeated 4 times ---

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...