NTA 4.4 Deadlocked Scheduler on SQL Server after upgrade from NTA 4.2

Since upgrading to NTA 4.4 and pointing NTA to a new database on our MS-SQL server about a week ago, the SQL Service has crash dumped twice with a Deadlocked Scheduler error.

We started over from scratch and did not attempt to migrate any data from 4.2 database. The SQL server is a beast and rarely breaks 25% CPU and 30% RAM capacity and was stable until it started hosting the NTA database.

I attempted to debug, but the server is in a Management VLAN without access to the Internet to download debug symbols and I am a Linux sysadmin with limited knowledge of MS-SQL.

While I am waiting for the Windows Database team to respond, Is there anything else that I should be looking at?

MS-SQL Server 2016 SP1 Enterprise

Model: Cisco UCSC-C240-M4SX

CPU: 2x Intel(R) Xeon(R) CPU E5-2623 v4 @ 2.60GHz 4 cores (8 threads)

RAM: 256 GB 2400 MHz DDR3

Disk: 24x 600GB Ent Performance SSD 3X in RAID10 (local, physical)

OS: Windows Server 2012 R2 64-bit

SQLDUMPER_ERRORLOG.log

Find more posts tagged with

nta 4.4

sql 2016

Accepted answers

All comments

oldefortran

The problem seems to involve multiple ESENT processes trying to access C:\Windows\system32\LogFiles\Sum\Api.log simultaneously possibly related to this article.

https://support.microsoft.com/en-us/help/2811566/error-1032-messages-in-the-application-log-in-windows-server-2012

oldefortran

It looks like stopping and disabling the User Access Logging Service (UAL) on the SQL server and all of the SolarWinds servers has resolved the problem. UAL is enabled by default on Windows Server >= 2012.

We are not using SQL Server Reporting Services and do not have AppInsight for SQL configured. I speculate that it could be triggered by some sort of database maintenance, but not sure.