I was receiving event ID 37, 50 from source dmio. The system is a HP Proliant DL380G3 with a SmartStart6400 Scsi raid card. I have sent to HP several log files but they could not find any problem in the hardware. I also applied the last firmware, reinstalled Windows 2000 and it did not work. I deleted the logical arrays from the default 5i and 6400 array controllers and recreated them again. Now it seems OK for I have not received any errors from the server.
I work with IBM xSeries servers and I want to add important information about the bad stripes that Kevin Skinner mentions. Bad stripe is a logical defect, so the only way to eliminate it, is by dismounting the Raid, creating a new one, and initializing the drives. Of course, you must have a backup first. Complete information about bad stripes can be found in the article “Troubleshooting ServeRAID bad stripe table entries”.
The event ID's 37, 50, and 30 have been showing up with a dmio source for about a week now on our IBM Netfinity 5100 with an EXP300 external SCSI raid cage. A tech support call to IBM lead us to run the ServeRAID Mgr. software and send the log files to IBM. This revealed that we had two drives with errors. So we then were asked to run a second log file/dump file with a utility that IBM sent us and it showed we have 2 bad stripes on our Raid 5. We are getting two new drives sent to us because we have bad blocks that are the main cause of these errors (according to IBM). The bad stripes we are just going to have to live with since putting new drives in will not fix the stripes only the blocks. I just hope we do not have to get at the data that went on those stripes. Hope this helps anyone else with these errors.
A prudent move would be to replace the disk causing the errors. If this disk is on an Exchange server you may start to see the following errors appear: "The read operation will fail with error -1018. If this condition persists then please restore the database from a previous backup."