On one of my SQL Server SNMP service continuously took nearly 20% of CPU. Rather than restarting the service I used the opportunity to completely restart the server and apply the latest up-to-day. At the launch of Windows I got an error message indicating a crash snmp.exe. I first thought about a port problem but the errors in the Event Viewer does not match:

Event Type: Error
Event Source: Service Control Manager
Event Category: None
Event ID: 7034
Date: 20/10/2009
Time: 6:52:40 p.m.
User: N / A
Computer: SERVERSQL
Description:
The SNMP Service service terminated unexpectedly for the first time.

Finally I discovered that the problem was an extension of the SNMP service installed by my Adaptec RAID controller card. To disable this extension simply run Regedit, go to HKEY_LOCAL_MACHINE \ SYSTEM \ CurrentControlSet \ Services \ SNMP \ Parameters \ ExtensionAgents and delete the value containing the data aussnmp SOFTWARE \ Adaptec \ RAIDController \ AAC \ CurrentVersion.

This figure is apparently he recreated every time you restart the machine will therefore be thinking of removing the value manually, or simply create a startup script to delete this value automatically.

Tags: Monitors , RAID , SNMP

On one of my servers Windows 2003 SNMP service could no longer boot properly: each attempt to start a rather uninformative message displayed:

The SNMP Service service on Local Computer started then stopped. Some services may stop automatically if they have no work to do, for example, the service alerts and logs performance.

In the event log, no error message but a series of informational message:

Event Type: Information
Event Source: Service Control Manager
Event Category: None
Event ID: 7035
Description: A start control has been correctly sent to the service SNMP Service.

Event type: Information
Event Source: Service Control Manager
Event Category: None
Event ID: 7036
Description: The SNMP Service service entered the state: stopped.

Event Type: Information
Event Source: SNMP
Event Category: None
Event ID: 1003
Description: The SNMP service was stopped successfully.

Nothing very clear about the source of the problem so ...

In my case I eventually discovered that the port 161 used by default by the SNMP service was used by another service. To solve the problem, just go to C: \ Windows \ system32 \ drivers \ etc and edit the services file. It must then change the line:

snmp              161/udp                           #SNMP

by

snmp              165/udp                           #SNMP

example.

Once this line has changed and saved the file, the SNMP service is properly started.

Tags: Monitoring , SNMP , Windows 2003

Cacti is a monitoring system of machines to monitor servers and workstations using the SNMP protocol in Perl scripts or using WMI on Windows. This system allows you to create graphics on various elements of a position such as CPU usage, statistics I / O on the disks or the types of queries on an IIS server.

In this article we will see how to install Cacti on a Windows 2003 + IIS 6 + PHP 4.4.9 + MySQL 5.1.30 + phpMyAdmin 2.11.9.4. To learn how to set up such a system, refer to this article .

Continue reading "Installing Cacti on Windows / IIS 6"

Tags: Cacti , Monitoring , Windows 2003

To determine which users who monopolize a SQL Server you can use the following query:

  SELECT
 TOP 20
 SPID Blocked,
 convert (varchar (10), db_name (dbid)) as Base
 CPU
 datediff (second, login_time, getdate ()) / 60 as Minutes,
 convert (float, cpu / datediff (second, login_time, getdate ())) as PScore,
 convert (varchar (16), hostname) as Host,
 convert (varchar (20), loginame) as Login,
 convert (varchar (50), program_name) as Program
 FROM master .. sysprocesses
 WHERE datediff (second, login_time, getdate ())> 0 and spid> 50
 ORDER BY desc PScore 

Over the PScore the higher the use of the database is intensive.

Similarly we can determine the most intensive programs observing BFactor column when running the following query:

  SELECT
 convert (varchar (50), program_name) as Program
 count (*) as CliCount,
 sum (cpu) have CPUSum,
 sum (datediff (Second, login_time, getdate ())) as SecSum,
 convert (float, sum (cpu)) / convert (float, sum (datediff (second, login_time, getdate ()))) as Score,
 convert (float, sum (cpu)) / convert (float, sum (datediff (second, login_time, getdate ()))) / count (*) as BFactor
 FROM master .. sysprocesses
 WHERE spid> 50
 GROUP BY
 CONVERT (varchar (50), program_name)
 ORDER BY score DESC 

Tags: Monitoring , SQL Server