November 24, 2010 at 6:47 am
I new to this so I found this quite fun!
This morning I got knee deep in reading about perfmon and profiler. After spending 20mins just trying to figure out how to access/open perfmon thinking it was a built in feature of Management Studio or somewhere in C:\MS SQL SERVER 2005\*. I discovered all I needed to do was run perfmon.msc and up it came. I have since setup some counters which are logining to C:\PerfLogs and I am watching it grow as I type! This is pretty cool! Well my question is this... once I start digging through this CSV file in Excel which i just did there are a Sh!ton of columns where each row represents a specific time for each measured column metric...How will I know what a bad reading is for any of these? For example SQL Cache Memory (KB) has several 1072 and 1088 entries. Is this good? Is this bad?
November 24, 2010 at 6:54 am
For example SQL Cache Memory (KB) has several 1072 and 1088 entries. Is this good? Is this bad?
It depends. What's normal for your system?
That's the answer for most of the 'is this bad' questions. That's why you need to have a baseline of what is normal so that you can easily tell what's abnormal.
Most of the counters are of minimal interest unless you're debugging very specific problems.
http://sqlinthewild.co.za/index.php/2009/02/14/do-you-know-how-to-use-perfmon/
Gail Shaw
Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability
November 24, 2010 at 8:31 am
GilaMonster (11/24/2010)
For example SQL Cache Memory (KB) has several 1072 and 1088 entries. Is this good? Is this bad?
It depends. What's normal for your system?
That's the answer for most of the 'is this bad' questions. That's why you need to have a baseline of what is normal so that you can easily tell what's abnormal.
Most of the counters are of minimal interest unless you're debugging very specific problems.
http://sqlinthewild.co.za/index.php/2009/02/14/do-you-know-how-to-use-perfmon/
Thanks Gila, I figured I may get such an answer. Baselining seems pretty important so I am glad I stumbled upon this when I did...starting out at a new career/company. Do you have any other suggestions on things I should look into knowing where I stand at this point in time?
November 24, 2010 at 9:17 am
Well I don't know where you stand and I don't know what your job description is so.... everything. 😉
If you don't have a firm grip of backups and restores, start there.
Is performance tuning your job? Indexing? SQL development? Do you need HA? What's your DR requirements and do you know how to fulfil them?
etc...
Gail Shaw
Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability
November 24, 2010 at 9:50 am
I'll add one thing, Profiler is great, but don't run it directly against the production systems in the company. On those look up how to do a server-side trace.
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
November 24, 2010 at 11:20 am
GilaMonster (11/24/2010)
Well I don't know where you stand and I don't know what your job description is so.... everything. 😉If you don't have a firm grip of backups and restores, start there.
Is performance tuning your job? Indexing? SQL development? Do you need HA? What's your DR requirements and do you know how to fulfil them?
etc...
Well long story short I am a recent college grad with very light experience with mySQL a few years back. The company I am now working for was primarly an Oracel shop but had a few hundred SQL Server DB's floating around. Some are supporting very important applications while most are not. Either way they never had a dedicated person to manage them. After working here a few weeks now it seems that rumors are that we will be moving more towards MS and away from Oracle. The problem is I am the only one of my kind and the Oracle DBAs are...well...um...ya. I will be going to a one week GK course in December which I am looking forward to. As far as high availability, yes! We have a few clustered environments. Backups were all setup on install and never touched again until I get around to them. Indexing is something that I want to do/learn more about. I will be expected to performance tune down the road and be able to do restores. I DO NOT have to worry about any development thankfully. So my role is to b/c the first SQL DBA and define that role while eventually bring others on board down the road as needed.
November 24, 2010 at 1:01 pm
Backups were all setup on install and never touched again until I get around to them.
Do you know if the backups are restorable in case of a disaster?
Indexing is something that I want to do/learn more about.
Go read my 3 indexing articles. If you're still after more info, go read the indexing category on my blog.
Also
http://sqlskills.com/BLOGS/PAUL/post/Important-things-for-DBAs-to-consider.aspx
http://technet.microsoft.com/en-us/magazine/gg299551.aspx
Gail Shaw
Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability
Viewing 7 posts - 1 through 6 (of 6 total)
You must be logged in to reply to this topic. Login to reply