How does SQL Response work...specifically?

dan_hdan_h Posts: 6
edited December 17, 2008 2:17PM in SQL Monitor Previous Versions
Hi, I have been reading up on SQL Response and that it is an agent-less monitoring tool and has a low overhead. But before I put it on one of my servers to test out, I would like to know how exactly it works.

Since it is a service installed on the repository server I presume it is making SQL calls on a periodic basis to the monitored servers (if this is incorrect please advise if it uses another means of collection (eg. memory scraping):

What calls is it making (dmv's, dbcc's, sys.db queries...etc...)?
How often does it make calls?
How many live/active enterprise servers have you monitored simultaneously in your testing?

Thanks,
--Dan

Comments

  • Hi Dan,
    Yes you are right the Alert Repository Service generates SQL and WMI queries periodically to each server in turn. (The sample rate can be modified via an XML configuration file.) We are constantantly looking at the SQL & WMI queries that we generate to ensure that they have low impact on the SQL Servers that are being monitored. (So for example for multiple SQL instances on one machine we will only request data such as Hard disk space once. You can use SQL Profiler to see what we are doing.)
    We have tested upto 80 SQL Servers here. (Most of these SQL Instances were running simulation SQL loads.)
    However if you do run into performance issues I would recommend that you install multiple Alert Repository services and then switch the UI between the different Alert Repositories.

    I hope that helps.
    Regards
    David
  • Thanks, I have installed it and am testing in my DEV environment.

    --Dan
Sign In or Register to comment.