I’m running a node failure test with I/O’s running under a Windows Cluster
Server 2003 environment and get the following Delayed Write Failure (if
this posting belongs in another newsgroup I apologize in advance):
Event Type: Warning
Event Source: Ntfs
Event Category: None
Event ID: 50
Date: 12/6/2004
Time: 1:27:33 AM
User: N/A
Computer: RAPHAEL
Description:
{Delayed Write Failed} Windows was unable to save all the data for the file
. The data has been lost. This error may be caused by a failure of your
computer hardware or network connection. Please try to save this file
elsewhere.
For more information, see Help and Support Center at
http://go.microsoft.com/fwlink/events.asp.
Data:
0000: 04 00 04 00 02 00 52 00 …R.
0008: 00 00 00 00 32 00 04 80 …2…€
0010: 00 00 00 00 10 00 00 80 …€
0018: 00 00 00 00 00 00 00 00 …
0020: 00 00 00 00 00 00 00 00 …
0028: 10 00 00 80 …€
The reason is 0x80040032 => IO_LOST_DELAYED_WRITE, and the error code
returned was 0x80000010 => STATUS_DEVICE_OFF_LINE. Sixteen seconds prior
to receiving this message the Cluster package sent a number of BusReset
requests to \Device\ClusDisk0 (I thought Cluster sent lun-level resets
first, followed by bus resets. Why it didn’t here I don’t know). Approx.
60 seconds prior to the failure one of two nodes in the Cluster was being
rebooted. There are also some errors from ‘ftdisk’ about transaction log
failures but one of Microsoft’s KB articles says this is okay.
What I’m trying to figure out is why wasn’t the Delayed Write re-attempted.
How long does Windows wait before giving up? 30 seconds, one minute, more?
Did it immediately fail because a STATUS_DEVICE_OFF_LINE was returned?
Thanks,
Bob