On Mon, 20 Feb 2006, steve szmidt wrote:
> On Sunday 19 February 2006 16:21, Eben King wrote:
>> So my friend D brought me a dead USB HD. 160 GB, about half full (he
>> said). XP hemmed and hawed when the drive was plugged in and wouldn't show
>> its contents. So I popped it on my box (Linux, natch) and got 67 GiB of
>> files from it, in several big tar files. Yay Linux. Took over two days,
>> because every time there was an error, it stopped and waited (for what?
>
> It was having access problems. (Checksum errors.) So it would reread over and
> over.
Fair enough. But disks are fast. It must have been rereading thousands of
times. You'd think it'd give up and return an error before then.
>> Now I've got everything I can get, and it's time to put it back. NTFS
>> writing is unreliable and new, so I chose FAT (vfat). Had to make 3
>> primary + virtual partitions, as XP's formatter won't make a 128 GiB FAT
>> partition (W98's will, but I can't get the drive to show up in W98).
>
> Ah, still using w98... OK, that's a problem in itself. I'd strongly recommend
> moving up to W2K SP2.
Other computers do use XP (until recently some used W2K), but that on's a
Celeron 450. Rather underpowered.
> Now you have a far more reliable and faster system. NTFS smokes FAT and is
> not unreliable under windows.
Agreed.
> Hmm. My experience is that NTFS (under Linux file system driver) is working
> fine. It is not exactly new as that has been in progress for years now. It's
> being considered experimental but I have had zero problems over the last
> couple of years.
OK. I thought the write support had severe limitations, like it could only
replace files, and with one of the same length too.
>> Copying to vfat is VERY slow -- in 6.5 hours it's written 11.3 GiB, which
>> is about 500 KiB/s. That's from an uncompressed tar file. It's slower on
>> big files, pausing for many seconds in the write.
>
> BTW, It's commonly written as KB and GB.
They're different: K=10^3=1000, G=10^9=1000000000; Ki=2^10=1024,
Gi=2^12=1073741824. http://physics.nist.gov/cuu/Units/binary.html
>> I have half a mind to create a big file, format it ext[23], loopback
>> mount it, extract the tar files to it, export it via Samba, mount it from
>> an XP laptop over wifi, and restore it THAT way. I'm afraid that's more
>> trouble than it's worth.
Did that. Slow but more convenient.
>> Also, my load went way up:
>>
>> eben@pc:/mnt/temp2$ uptime
>> 16:15:31 up 3 days, 16:43, 9 users, load average: 10.70, 10.72, 10.74
>> How do I find out what those "waiting" processes are? Can I convince "ps"
>> to tell me?
>
> Waiting is normal, commands are queued. Executing a command like:
> ps -eo pid,tid,class,rtprio,ni,pri,psr,pcpu,stat,wchan:14,comm
>
> will tell you more of what is going on. Be prepared to use the man page!
Thanks. ps(1) is quite confusing.
eben@pc:~$ ps -eo pid,tid,class,rtprio,ni,pri,psr,pcpu,stat,wchan:14,comm | wc -l
128
I guess that shows processes waiting on _any_ syscall.
-- -eben ebQenW1@EtaRmpTabYayU.rIr.OcoPm home.tampabay.rr.com/hactarQ: What kind of modem did Jimi Hendrix use? A: A purple Hayes. ----------------------------------------------------------------------- This list is provided as an unmoderated internet service by Networked Knowledge Systems (NKS). Views and opinions expressed in messages posted are those of the author and do not necessarily reflect the official policy or position of NKS or any of its employees.
This archive was generated by hypermail 2.1.3 : Fri Aug 01 2014 - 18:35:56 EDT