SQL Error 833 - Long I/O Request since 9th Feb 2026 Defender Release of new SQL feature

Question

Hi, I'm seeing a lot of SQL Error 833 - Long I/O Requests on certain SQL Servers since around midnight last Sunday. I can't find any underlying infrastructure/ storage reason for this or any commonality, and it's only affecting a select few SQL Servers.

The only thing I can find is a Defender release around the same time of a new feature that allows simulation of alerts for SQL Server. It may be completely unrelated but is anyone else seeing this since 8/9th Feb?

Answer 1

The cause was a failing SFP as per below. It was nothing to do with Defender which we managed to rule out by running logs.

Answer 2

Message 833:

SQL Server has encountered %d occurrence(s) of I/O requests taking longer than %d seconds to complete on file [%ls] in database id %d. The OS file handle is 0x%p. The offset of the latest long I/O is: %#016I64x. The duration of the long I/O is: %I64u ms.

This suggest serious problems with your I/O subsystem. You say that you don't find any issue in underlying infrastructure, but I/O subsystems consist of many components.

On way to find out whether this may some simulation (which I doubt) is to log the output sys.dm_io_virtual_file_stats to a table every ten minute or so, and then compute deltas, and from deltas, you can compute average time for I/O requests. I would expect that you will find some pretty bad numbers.

Performance Monitor is also an option.

I assume that this is on-prem?

Answer 3

SQL Error 833 indicates that SQL Server has encountered I/O requests taking longer than 15 seconds to complete. This type of error often points to underlying issues with the I/O subsystem rather than SQL Server itself. Since you mentioned that this issue started occurring around the time of the Defender release, it is worth investigating whether there are any changes in the environment or workload that could correlate with this timing.

Here are some potential causes and steps to investigate:

1. I/O Latency: Check the Avg Disk Sec/Transfer counter for the affected SQL Servers. If the values are consistently above 10-15 milliseconds, further investigation into the I/O subsystem is warranted.
2. Workload Intensity: Determine if there has been an increase in workload on these SQL Servers that could be causing the long I/O requests. Heavy operations such as backups, large queries, or bulk inserts can exacerbate I/O issues.
3. Antivirus or Security Software: Ensure that any antivirus or security software is configured properly, as these can sometimes interfere with SQL Server operations and lead to increased I/O latency.
4. Performance Monitor Counters: Utilize Performance Monitor to check for long disk latencies, long disk queues, or any other indicators of I/O subsystem issues.
5. Error Log Review: Review the SQL Server error logs for any additional context or patterns related to the long I/O requests.

While the new Defender feature may not be directly related, it’s prudent to monitor the systems closely for any changes in performance or behavior following the release.