SD 2.6.1 stops randomly

[mhmertens]

Leo 10.5.8, SD 2.6.1
Smart upd to image on ext disk (FW800 or 400). 1.2 mio files in total.

SD stops for no clear reason. Last 2 commands from log after having evaluated approx 100 000 files.
| 02:26:26 PM | Info | Ignoring /Desktop DF
| 02:26:26 PM | Info | /Previous Systems.localized

Stopping means that even after an hour there is no progress at all.

Finder reactions are terribly slow in this state. (slow means I have to wait for 10-15 mins before something reacts)
SD reactions as well slow but just about bearable.

External drive encl (brand=Pleiades) has Oxford chipset (so should be the best there is)

Finder and SD appear in the force quit dialog as 'not responding' but I know they may well still be working on whatever.

This is a reproducible problem.

How do I go on to debug this further? Any help much appreciated.

Regards

mHm

[dnanian]

Likely that either your source or destination volume are experiencing severe problems, actually. Given that it's stopping in Previous Systems, I'd delete that folder, empty your trash and try again. If that resolves the issue, your internal drive is probably failing, and I'd consider replacing it.

[mhmertens]

Likely that either your source or destination volume are experiencing severe problems, actually. Given that it's stopping in Previous Systems, I'd delete that folder, empty your trash and try again. If that does not resolve the issue, your internal drive is probably failing, and I'd consider replacing it.

I've taken the liberty to correct what you said, surely a typo:)
I keep the "previous systems" folder because I need to refer to it when gradually (over weeks) need to get certain settings files etc out from there to put in the current OS. How can I at least back it up, before I delete it. Alternativly, can I instruct SD to exclude it from the copy/backup?

Pls note: Both the the external drive and the entire MacBook Pro are less than 3 months old. I know this does not guarantee that they cannot fail, but its unlikely.

[dnanian]

Actually, no: I meant "resolves the", because it would indicate that the source drive is causing the hang-up, and thus has bad sectors on it, never a good thing.

I'm kind of surprised that you might need Previous Systems for anything. The vast majority of configuration files are preserved when you do the archive-and-install or upgrade.

It doesn't really matter how old the drives are. One of them certainly seems to be failing. You might find it useful to open the Console application, put it on "All messages" and look for kernel messages regarding I/O errors on devices: that could point you to the troublesome disk (or whatever it might be), too...

[mhmertens]

Ok, I did not fully get what you meant originally.
I found indeed I/O, but not for the internal disk. disk0s1=internal, disk1s1= external, disk2s1 =image on external

08.10.09 14:26:03 SuperDuper![12214] .sdef error: Operation could not be completed. (NSXMLParserErrorDomain error 1549.)
08.10.09 14:26:03 SuperDuper![12214] line number: 2

08.10.09 14:26:03 SuperDuper![12214] .sdef error: Operation could not be completed. (NSXMLParserErrorDomain error 1549.)
08.10.09 14:26:03 SuperDuper![12214] line number: 2
08.10.09 14:26:03 SuperDuper![12214] .sdef error: Operation could not be completed. (NSXMLParserErrorDomain error 1549.)
08.10.09 14:26:03 SuperDuper![12214] line number: 2
08.10.09 14:26:03 [0x0-0x2a32a3].com.blacey.SuperDuper![12214] crontab: no crontab for mhm
08.10.09 14:26:15 KernelEventAgent[37] tid 00000000 received unknown event (256)
08.10.09 14:26:16 sudo[12233] root : TTY=unknown ; PWD=/ ; USER=mhm ; COMMAND=/bin/chmod 0775 /Volumes/Samsung 2/Leo HD 17 MBP 4Sep09.sparseimage
08.10.09 14:26:16 sudo[12235] root : TTY=unknown ; PWD=/ ; USER=mhm ; COMMAND=/usr/bin/hdiutil attach /Volumes/Samsung 2/Leo HD 17 MBP 4Sep09.sparseimage -noverify -noautofsck
08.10.09 14:26:19 hdiejectd[12241] running

08.10.09 14:26:19 kernel jnl: disk3s2: replay_journal: from: 4679168 to: 12281856 (joffset 0x4a8000)
08.10.09 14:26:21 kernel jnl: disk3s2: journal replay done.
08.10.09 14:26:21 fseventsd[41] event logs in /Volumes/Leo HD 17 MBP 4Sep09/.fseventsd out of sync with volume. destroying old logs. (78713 2 152253)
08.10.09 14:26:22 fseventsd[41] log dir: /Volumes/Leo HD 17 MBP 4Sep09/.fseventsd getting new uuid: F432051E-6D5C-49D2-8EC8-6D74E3B39544
08.10.09 14:26:23 KernelEventAgent[37] tid 00000000 received unknown event (256)
08.10.09 14:27:09 /usr/sbin/ocspd[12267] starting

08.10.09 14:32:07 kernel disk1s3: I/O error.
08.10.09 14:32:07 kernel disk3s2: 0x5 (UNDEFINED).
08.10.09 14:33:08 kernel jnl: disk1s3: flushing fs disk buffer returned 0x5
08.10.09 14:34:08 kernel disk1s2: I/O error.
08.10.09 14:35:39 kernel jnl: disk1s2: flushing fs disk buffer returned 0x5
08.10.09 14:39:39 kernel disk1s3: I/O error.
08.10.09 14:39:39 kernel disk3s2: 0x5 (UNDEFINED).
08.10.09 14:40:39 kernel disk1s3: I/O error.
08.10.09 14:40:39 kernel jnl: disk1s3: do_jnl_io: strategy err 0x5
08.10.09 14:40:39 kernel jnl: disk1s3: write_journal_header: error writing the journal header!
08.10.09 14:41:19 kernel IOSCSIPeripheralDeviceType00::setPowerState(0xb914 000, 3 -> 4) timed out after 100112 ms

[dnanian]

Ah, there you go. So, the problem is with the external drive...

[tdurocher]

After having this issue happen to me, and then googling for setpowerstate, I find that it has been fairly common for a couple years now. It has nothing to do with SD. Many people think it's a Time Machine problem but it is just that TM (and occassionally SD) are common programs that try to bring an external drive out of sleep. The problem is that the drives won't come out of sleep. So it appears to be a driver problem. Since it happens with multiple drive vendors, I'm guessing its the Apple side of the drivers. On the other hand, they don't seem to be fixing it.

The workaround is to go into Energy Saver preferences and turn off Put Hard Drives to Sleep. Not something I really feel good about doing. As for a fix, I have yet to find one....