October 15, 2008 at 11:47 pm
I have a dell poweredge 2950 with PERC 5 SCSI RAID 1. 4GB of memory. Everything is fine. Its a rock solid server. Its running Windows Server 2003 Enterprise 32 Bit with SP1
I want to install and use SQL 2000 or 2005 on it. So I run the stress
tester sqliosim.exe and I get like 1200 errors immediately.
For example (like 500 of these):
Error: 0x800705AA
Error Text: Insufficient system resources exist to complete the
requested service.
Description: Write IO Error
Another 500 or so of these:
Error: 0x80070467
Error Text: While accessing the hard disk, a disk operation failed
even after retries.
Description: Buffer validation failed on C:\sqliosim.mdx Page: 113398,
offset 0x0
No errors in the event viewer at all ever. Ther server is fine.
Any idea?? Theres plenty of space, just a hardware RAIDed C:\ and
again, its fine. Nothing is even running on it.
Thanks for any help.
October 16, 2008 at 10:04 am
This means that your I/O subsystem is causing corruptions. What does it say in the extended description for these errors?
Is this a new system? If not, have you ever seen I/O issues with it before? Are all the drivers and firmware up-to-date? Remember that SQLIOsim *really* pushes it beyond what its likely to experience in real life.
Paul Randal
CEO, SQLskills.com: Check out SQLskills online training!
Blog:www.SQLskills.com/blogs/paul Twitter: @PaulRandal
SQL MVP, Microsoft RD, Contributing Editor of TechNet Magazine
Author of DBCC CHECKDB/repair (and other Storage Engine) code of SQL Server 2005
October 16, 2008 at 7:13 pm
That is the extended desciption. Heres one again:
10/15/08 21:25:27 2316 CPU Idle User Error: 0x80070467
Error Text: While accessing the hard disk, a disk operation failed even after retries.
Description: Buffer validation failed on C:\sqliosim.mdx Page: 113600, offset 0x0 CBUF::ValidateBuffer e:\yukon\sosbranch\sqltdbms\storeng\util\sqliosim\buffer.cpp 791
Its a system that has been in operation for close to 2 years. No I/O problems at all ever. The firmware is mostly up to date. I'll verify that but still, this should come back fine. So many of the same errors. Its strange.
October 16, 2008 at 7:26 pm
Unfortunately there are no good resources available to really diagnose these issues in-depth. On all the customer systems I've seen, there has been a fault found when SQLIOsim reported errors.
All I can suggest is to call Product Support and see if they can help you further. I'll also ping some folks on the dev team to see if they know of anything that could cause spurious errors.
Thanks
Paul Randal
CEO, SQLskills.com: Check out SQLskills online training!
Blog:www.SQLskills.com/blogs/paul Twitter: @PaulRandal
SQL MVP, Microsoft RD, Contributing Editor of TechNet Magazine
Author of DBCC CHECKDB/repair (and other Storage Engine) code of SQL Server 2005
October 16, 2008 at 9:48 pm
This particular message is logged from a routine that loads a random page from the test file and checks that the data retrieved is the data expected. With the error log there should be some additional files in the errorlog directory that contain the full page dumps of the actual and expected data. You should be able to see a difference between the two pages there.
Assuming the output does show different data, then even though your system may appear to be behaving correctly this failure would lead me to believe that there is a deeper problem on the system.
Jerome Halmans
Developer, SQLOS
Viewing 5 posts - 1 through 4 (of 4 total)
You must be logged in to reply to this topic. Login to reply