Why does opensuse 11.3 take so long to load? It hangs for minutes at beg!

I just recently installed openSUSE 11.3 (clean, on an empty HD) after previously having 11.2 on here (but wiping that out). When I start up 11.3, it hangs for 3+ minutes at the first screen with the lizard before finally moving the white line! This never happened before. It now takes 4-5 times as long to start up as with 11.2.

Any ideas?

Press esc to see the boot process maybe you can see what is taking so long.

OK, I saw a bunch of messages but I wasn’t able to copy it all down. Where is the log from startup located?

Here’s what I saw:


exception Emask 0x0 Sact 0x0 SErr action 0x6 frozen
failed command write DMA
emask 0x4 timeout

OK, here’s the same error messages from /var/log/messages:


linux kernel:   447.000044] ata1.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen
linux kernel:   447.000050] ata1.00: failed command: WRITE DMA EXT
linux kernel:   570.000117]          res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)

This is obviously a HD failure. Check all connections and test the drive itself for bad blocks etc. (for example by using s.m.a.r.t. or fsck).

On 2010-09-21 17:08, 6tr6tr wrote:
>
> OK, here’s the same error messages from /var/log/messages:
>
>
> Code:
> --------------------
>
> linux kernel: 447.000044] ata1.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen
> linux kernel: 447.000050] ata1.00: failed command: WRITE DMA EXT
> linux kernel: 570.000117] res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
>
> --------------------

This same disk had 11.2 with no provblem, previously? Then install both systems on different
partitions, report this in a bugzilla, and use 11.2 till it is solved.

It is a different, new, disk? Test the disk for problems with the manufacturer supplied diagnostic
utility, and return disk if it does not pass.


Cheers / Saludos,

Carlos E. R.
(from 11.2 x86_64 “Emerald” GM (Elessar))

There is no utilitity that came with the disk. I’ll check online but usually those only work in Windows. Is there some way I can check it with smartctl or will that not find anything?

You could try testing it with “smartctl”. Type “man smartctl” in a console for detailed instructions.

Both --health and -a say that it passed with no errors (for both disks: sda, sdb).

Does it finally boot completely or is the system unstable? If it boots completely, with all partitions mounted, I suggest you first perform an online update, then reboot, get back here with results.

BTW. Do you have NFS client enabled ?

This is what I think is happening. You have one new fast disk (transfer speed) and one slow old one. On a single cable the transfer speed will be reduced to the slowest device. So the system sees fast disk first and all is well but when we get to the second one whoops can’t speak that fast. So we drop the speed and try again again whoops can’t talk until at last we reach an acceptable speed.

Try putting the slow device on another cable.

On 2010-09-23 17:06, 6tr6tr wrote:
>
> Carlos E. R.;2226874 Wrote:

> There is no utilitity that came with the disk. I’ll check online but
> usually those only work in Windows. Is there some way I can check it
> with smartctl or will that not find anything?

I never said that the utility would come with the disk. The manufacturers usually supply them
online, as a bootable floppy or CD. Bootable means that no operating system of yours is required.

Yes, smartctl is a posibility, but it will not test cables or cpu. Notice that -a or health is not
enough, you need the full long test at least - and it will not test the cable.

Use the manufacturer utility this time.


Cheers / Saludos,

Carlos E. R.
(from 11.2 x86_64 “Emerald” GM (Minas Tirith))