Re: Linux1394-user digest, Vol 1 #1000 - 12 msgs

In a message dated 2/10/2004 8:25:53 PM Eastern Standard Time,
linux1394-user-request@lists.sourceforge.net writes:
> Details:
> Kernel: 2.4.22 (no ieee1394 patches)

Bad choice. sbp2 was unstable and vfat driver can destroy file contents
on large filesystems (hiding free space).

Eduard.

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

Can you tell me what you mean that Kernel 2.4.22 was a "bad choice" for sbp2?
I am running 2 5-disk RAID5 arrays using firewire drives, 5 individual PCI
firewire host adapters, and Mandrake 9.2 (kernel = mandrake enterprise
2.4.22-21). I'm getting great performance in general. My only complaints are

1) I'm getting a few errors occasionally having to do with reconnection to
SPB-2 device. I'll see the following in my Kernel errors log:
Feb 9 09:12:42 localhost kernel: ieee 1394: spb2: Node 0-00:1023: Max speed
[S400] - Max payload [2048]
Feb 9 09:12:42 localhost kernel: ieee 1394: spb2: Node 0-02:1023: Max speed
[S400] - Max payload [2048
Feb 9 09:12:42 localhost kernel: ieee 1394: spb2: Node 1-00:1023: Max speed
[S400] - Max payload [2048
Feb 9 09:12:42 localhost kernel: ieee 1394: spb2: Node 1-01:1023: Max speed
[S400] - Max payload [2048
Feb 9 09:12:42 localhost kernel: ieee 1394: spb2: Node 2-00:1023: Max speed
[S400] - Max payload [2048
Feb 9 09:12:42 localhost kernel: ieee 1394: spb2: Node 2-01:1023: Max speed
[S400] - Max payload [2048
Feb 9 09:12:42 localhost kernel: ieee 1394: spb2: Node 3-00:1023: Max speed
[S400] - Max payload [2048
Feb 9 09:12:42 localhost kernel: ieee 1394: spb2: Node 4-02:1023: Max speed
[S400] - Max payload [2048
Feb 9 09:12:42 localhost kernel: ieee 1394: spb2: Error reconnecting to SPB-2
device - reconnect failed
Feb 9 09:12:42 localhost kernel: ieee 1394: spb2: Node 0-00:1023: Max speed
[S400] - Max payload [2048
Feb 9 09:12:42 localhost kernel: ieee 1394: spb2: Error reconnecting to SPB-2
device - reconnect failed
Feb 9 09:12:42 localhost kernel: ieee 1394: spb2: Node 0-01:1023: Max speed
[S400] - Max payload [2048
Feb 9 09:12:42 localhost kernel: ieee 1394: spb2: Node 4-00:1023: Max speed
[S400] - Max payload [2048
Feb 9 09:12:42 localhost kernel: ieee 1394: spb2: Node 4-01:1023: Max speed
[S400] - Max payload [2048

Sometimes I'll get more errors -- always in pairs -- but usually it's just
one pair.
Don't know if I should worry. IS THIS the kind of thing you mean by
"instability" under 2.4.22?

Also, I've had a few incidents in which the activity lights on one or the two
drives that's plugged into the FIRST of my 5 host adapters lit up and stayed
lit, thus freezing the RAID array. The only solution was to force a shutdown.
The drives that were on that Host adapter got marked as "faulty" and were
removed from the RAID arrays, forcing me to run a resync operation on the arrays
(fortunately no data lost).

I just tried replacing the Host Adapter card -- as I have had the same thing
happen to three different drives and they were all plugged into the same Host
Adapter card (2nd PCI slot from my AGP graphics cards). I thought maybe the
host adapter was faulty.

All my host adapters are based on TI-chips, 4 used to show up as
"TSB12LV26 OHCI-Lyncx PCI IEEE 1394 Host Controller" in Mandrake's Hardrake2
program. One showed up as "
TSB43AB23 IEEE-1394 Controller (PHY-Link) 1394a-2000".
The drives giving me problems were plugged into one of the "TSB12LV26" cards
so I replaced the card and now have 2 TSB43AB23 cards and 3 TSB12LV26 cards.

I don't know if it's a bad idea to mix the types of chips on the cards. I
called Texas Instruments to find out if one chip or the other was better for
SBP-2 applications but they told me they couldn't comment on that!

Of course, my problem could also have to do with IRQ sharing. Given that my
RAID arrays currently have data on them, it's almost impossible for me to
figure out which card is assigned to which IRQ, and whether the host adapter that's
giving me problems is sharing its IRQ with too many other devices, or the
wrong devices. (For instance, my Gigabit ethernet card is carrying a ton of data
to and from my Linux box. My drive freeze-ups have ALWAYS occurred under
intense network loads.

But could these problems be related to the 2.4.22 kernel????? If so, why? And
what kernel fixes the problem???

Thanks for your replies.

-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration
See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn
_______________________________________________
mailing list Linux1394-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux1394-user
Received on Wed Feb 11 15:16:20 2004