I keep getting a disk warning 51 message in the system log every few months. It eventually causes the server to crash (killing all my network services), forcing me to do a hard reboot.

Full error (repeats until the server is rebooted):

An error was detected on device \device\harddisk0\dr0 during a paging operation.

I also get (error 6): An I/O operation initiated by the Registry failed unrecoverably. The Registry could not flush hive: '\??\c:\users\someuser\ntuser.dat'.

I'm running some network/disk intensive applications: one is visual studio 2005 mvc application that collects a data feed, writes some of it to a log file and forwards the whole feed, via a socket, to a java application.

The java 6 application also writes most of the data to a log and some of it to a mysql 5.1 database. I have java 6 update 20 (gonna try updating this).

I know this is a lot of redundant data, but I like to have it around for debugging purposes.

I'm thinking that the network buffers are getting overloaded. This is a Dell Poweredge R710 with a broadcom netxtreme II bcm5709c nic. May try getting another card.

In the past I tried installing the latest windows updates and updating my Dell SAS 6/iR controller driver.

Anyone have an issue like this? Any help is appreciated.

  • it looks like a hard drive may be failing. How are the disk drives set up on this machine?
    – Lamar B
    Dec 7, 2011 at 2:34
  • That was my initial suspicion, but this server is pretty new (only had a year and a half) and the first time I got this error was 3 months after it was installed. Ran dell diagnostics on disk and checked out ok. I only have one drive, which is partitioned into a system and data section. The drive has write caching enabled.
    – James
    Dec 7, 2011 at 3:20
  • A] make a backup now if you haven't already just to be safe B] do you have a way to check the SMART data coming from the drive?
    – Lamar B
    Dec 7, 2011 at 3:23
  • I'm in the process of installing open manage on there. I know in the past some dell set ups I worked on gave the smart status in the windows event log. I will see about getting a smart tool on there. Thanks for the pointer.
    – James
    Dec 7, 2011 at 3:33
  • The 5709 chips had an updated firmware released last Dec that fixed a lot of problems; be sure you've updated since then. What kind of RAID, HBA, disk subsystems, etc are you using? Are they updated and have all the management software installed? Any reports in the management software? Are you using officially supported Disks in the server?
    – Chris S
    Dec 7, 2011 at 3:35

2 Answers 2


Not to repeat Lamam B but I have done this recently.

Do a backup, ShadowProtect Server is great if you want to mirror the whole thing and drop it onto a new hard disk for minimal down time. I did it a couple of months ago and was back online within 2 hours.

I am sure there is alternatives out there but SP is pretty good.

  • Thanks. I was doing backups sporadically. Now I have an automated backup, that copies stuff to the cloud, in case anything happens.
    – James
    Feb 14, 2012 at 18:10

Chris' suggestion did the trick.

I updated the firmware for broadcom netxtreme II bcm5709c nic (I was reluctant to do at first this since I don't have physical access to the server and didn't want to lose my network connection).

We haven't had any crashes since then.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.