Re: Filesystem problems at boot and shutdown?

2008-11-27 Thread Wojciech Puchar

Nov 26 20:48:11 server kernel: ad4: WARNING - WRITE_DMA48 UDMA ICRC error
(retrying request) LBA=1364750271
Nov 26 20:48:11 server kernel: ad4: FAILURE - WRITE_DMA48
status=51 error=4 LBA=1364750271
Nov 26 20:48:11 server kernel:
g_vfs_done():ad4s1g[WRITE(offset=682932174848, length=131072)]error = 5



cabling problems or disk overheating (or controller) or failed disk.

ICRC error suggest that it's rather cabling problem.

___
freebsd-questions@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-questions
To unsubscribe, send any mail to "[EMAIL PROTECTED]"


Re: Filesystem problems at boot and shutdown?

2008-11-27 Thread Manolis Kiagias
Rick Janssen wrote:
>> Rick Janssen wrote:
>> 
>>> I've been playing around with FreeBSD for some time now, still being
>>> unable to solve some problems. Let me explain.
>>>
>>> I'm trying to run a webserver on the machine. Just basic, nothing too
>>> fancy. Problem concerns the following: The website served is speedy as
>>> expected when accessed from local LAN, but when accessed from WAN via
>>> router, it's realy slow. So, to do some tests, this afternoon I requested
>>> some pages from an computer outside my LAN. Server was very slow again,
>>> even ssh slowed to a crawl. Suddenly, without reason I know off,
>>> everything sped up. I issued a reboot to check if the problem might have
>>> been 'solved'. This took a long time.
>>>
>>> Back home I checked the logs. It now appeared the long reboot-time was
>>> caused by a Syncing disk anomaly, which happens when the system prepares
>>> for shutting down.
>>>
>>> Nov 26 22:38:12 server kernel: Syncing disks, vnodes remaining...1 1 1 1
>>> 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
>>> 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 timed out
>>> Nov 26 22:38:12 server kernel: Waiting (max 60 seconds) for system process
>>> `bufdaemon' to stop...done
>>> Nov 26 22:38:12 server kernel: Syncing disks, buffers remaining... 184 185
>>> 186 1 185 185 185 185 184 184 185 185 185 184 184 185 1 184 184 184 184
>>> 185 185 185 1 184 184 184 184 184 185 184 184 185 185 1 185 184 184 184
>>> 184 184 184 1 185 184 184 184 184 185 184 184 185 1 185 184 184 184 184
>>> 184 184 185 1 184 184 185 184 184 184 184 185 184 184 185 1 185 185 184
>>> 184 184 185 184 184 184 185 1 184 184 185 184 184 185 184 184 185 185 184
>>> 184 184 184 184 1 184 184 184 1 184 184 185 1 185 184 184 1 184 185 184
>>> 184 185 185 184 184 ... etc etc...
>>> Nov 26 22:38:12 server kernel: Final sync complete
>>>
>>> So I figured, might as well run fschk -y in single user mode to fix
>>> potential problems. Now I got some new errors:
>>>
>>> Nov 26 20:48:11 server kernel: ad4: WARNING - WRITE_DMA48 UDMA ICRC error
>>> (retrying request) LBA=1364750271
>>> Nov 26 20:48:11 server kernel: ad4: FAILURE - WRITE_DMA48
>>> status=51 error=4 LBA=1364750271
>>> Nov 26 20:48:11 server kernel:
>>> g_vfs_done():ad4s1g[WRITE(offset=682932174848, length=131072)]error = 5
>>>
>>> So far, nothing serious has followed from these errors I know off. They
>>> only happen sporadically, during reboots.
>>>
>>> Are these to problems even related, or am I just unlucky? Anyone has some
>>> suggestions to fix them?
>>>
>>> Regards,
>>> Rick
>>>
>>>
>>>   
>> Unless there is some incompatibility between FreeBSD -> Your disk
>> controller -> Your disk, my best guess is you have a failing disk.  I
>> would also suggest you check cables, connections (I guess this is an
>> ATA disk so you may wish the check whether the flat cable is the
>> 80-conductor type and is plugged in correctly).
>> 
>
> The disk is a new SATA model (1TB Samsung), so that lessens the chances of
> the disk itself being defective somewhat. The mainboard is a MSI K9A2VM-F
> (AMD 780V / AMD SB700 chipsets)
>
>   

In fact, new disks *can* be faulty. Usually, the faults surface few days
/ weeks after initial use. I don't really trust a new disk on my system
until it runs for a couple of months. It will then run for years until
it reaches end-of-life due to wear.
I still suggest you try testing it with some programs like
sysutils/smartmontools

>> You did not mention the speed of you WAN connection, but FWIW I am
>> running a webserver behind a 1Mbps/256Kbps ADSL line and response is
>> good enough. SSH is definitely good enough to use in long vi sessions,
>> with lengthy documents.
>> 
>
> The connection is something similar. Should be no cause for the slowness.
>   

Since it works fine on the LAN, you need to check the router though.
Maybe some setting is slowing it down. Is it one of these cheap DSL
routers? You may want to turn off various features on the router like
firewalls, intrusion detections and so on (it goes without saying that
you will secure your FreeBSD system instead)
___
freebsd-questions@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-questions
To unsubscribe, send any mail to "[EMAIL PROTECTED]"


Re: Filesystem problems at boot and shutdown?

2008-11-26 Thread Manolis Kiagias
Rick Janssen wrote:
> I've been playing around with FreeBSD for some time now, still being
> unable to solve some problems. Let me explain.
>
> I'm trying to run a webserver on the machine. Just basic, nothing too
> fancy. Problem concerns the following: The website served is speedy as
> expected when accessed from local LAN, but when accessed from WAN via
> router, it's realy slow. So, to do some tests, this afternoon I requested
> some pages from an computer outside my LAN. Server was very slow again,
> even ssh slowed to a crawl. Suddenly, without reason I know off,
> everything sped up. I issued a reboot to check if the problem might have
> been 'solved'. This took a long time.
>
> Back home I checked the logs. It now appeared the long reboot-time was
> caused by a Syncing disk anomaly, which happens when the system prepares
> for shutting down.
>
> Nov 26 22:38:12 server kernel: Syncing disks, vnodes remaining...1 1 1 1 1
> 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
> 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 timed out
> Nov 26 22:38:12 server kernel: Waiting (max 60 seconds) for system process
> `bufdaemon' to stop...done
> Nov 26 22:38:12 server kernel: Syncing disks, buffers remaining... 184 185
> 186 1 185 185 185 185 184 184 185 185 185 184 184 185 1 184 184 184 184
> 185 185 185 1 184 184 184 184 184 185 184 184 185 185 1 185 184 184 184
> 184 184 184 1 185 184 184 184 184 185 184 184 185 1 185 184 184 184 184
> 184 184 185 1 184 184 185 184 184 184 184 185 184 184 185 1 185 185 184
> 184 184 185 184 184 184 185 1 184 184 185 184 184 185 184 184 185 185 184
> 184 184 184 184 1 184 184 184 1 184 184 185 1 185 184 184 1 184 185 184
> 184 185 185 184 184 ... etc etc...
> Nov 26 22:38:12 server kernel: Final sync complete
>
> So I figured, might as well run fschk -y in single user mode to fix
> potential problems. Now I got some new errors:
>
> Nov 26 20:48:11 server kernel: ad4: WARNING - WRITE_DMA48 UDMA ICRC error
> (retrying request) LBA=1364750271
> Nov 26 20:48:11 server kernel: ad4: FAILURE - WRITE_DMA48
> status=51 error=4 LBA=1364750271
> Nov 26 20:48:11 server kernel:
> g_vfs_done():ad4s1g[WRITE(offset=682932174848, length=131072)]error = 5
>
> So far, nothing serious has followed from these errors I know off. They
> only happen sporadically, during reboots.
>
> Are these to problems even related, or am I just unlucky? Anyone has some
> suggestions to fix them?
>
> Regards,
> Rick
>
>   
Unless there is some incompatibility between FreeBSD -> Your disk
controller -> Your disk, my best guess is you have a failing disk.  I
would also suggest you check cables, connections (I guess this is an 
ATA disk so you may wish the check whether the flat cable is the
80-conductor type and is plugged in correctly). Hopefully, if you are
just playing with the system, you don't have any critical data in there,
otherwise I would suggest you back up immediately.  Do you have a spare
disk to try and see what happens?
Whether your other problem depends on this: It could be, since your
webserver might be trying to read from a faulty area and keep retrying. 
Or indeed there is some incompatibility and the disk is constantly
under-performing. 
You did not mention the speed of you WAN connection, but FWIW I am
running a webserver behind a 1Mbps/256Kbps ADSL line and response is
good enough. SSH is definitely good enough to use in long vi sessions,
with lengthy documents.
___
freebsd-questions@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-questions
To unsubscribe, send any mail to "[EMAIL PROTECTED]"