Press "Enter" to skip to content

Author: Kevin Feasel

When Secondary Nodes Affect Primaries

Published 2018-09-13 by Kevin Feasel

Dmitri Korotkevitch shows that a readable secondary in an Availability Group can negatively impact the primary:

A long time ago in a galaxy far, far away, I had to troubleshoot interesting performance issue in SQL Server. Suddenly, the CPU load on the server started to climb up. Nothing changed in terms of workload. The system still processed the same amount of requests. The execution plans of the critical queries stayed the same. Nevertheless, the CPU usage grew up slowly and steadily by a few percent per hour.

Eventually, we nailed it down. The problem occured in very busy OLTP system with very volatile data. We noticed that system performed much more I/O (logical and physical) than before. It was very strange, because nothing should have changed that day. Finally, we found that we have large number of deleted rows in the database that had not been cleaned up by ghost cleanup task.

Read on to learn what caused this mess.

Comments closed

Issues With Bulk Inserting Multi-Byte Characters In Fixed Width Files

Published 2018-09-13 by Kevin Feasel

Randolph West shares an example of an issue with BULK INSERT:

Fellow Canadian Doran Douglas brought this issue to my attention recently, and I wanted to share it with you as well.

Let’s say you have a file in UTF-8 format. What this means is that some of the characters will be single-byte, and some may be more than that.

Where this becomes problematic is that a fixed-width file has fields that are, well, fixed in size. If a Unicode character requires more than one byte, it’s going to cry havoc and let slip the dogs of truncation.

Click through for an example.  This seems like a bug to me—I interpret fixed-width as fixed number of characters, not fixed number of bytes.  At the very least, it’s liable to cause confusion.

Comments closed

Trigger Spirals

Published 2018-09-13 by Kevin Feasel

David Fowler tells a story of woe, one which is totally not his fault:

To do this, a trigger was created which would send all the details via a Service Broker message to another SQL Server, this SQL Server was used to hold details of the AD accounts and from there, changes were automatically propagated out to AD.

This was working well until one day when it was realised that any changes to account permissions in AD weren’t reflected in the personnel database.  To solve this, another trigger was created to send a Service Broker message back to the personnel database with details of the change.

This was where I came in, it was noticed that the system had started to run slower and slower, not only that but permissions seemed to be constantly changing for no obvious reason.  Were the machines finally waking up and taking over?

There’s a reasonable explanation here, for some definition of reasonable.

Comments closed

Subsetting Lists In R

Published 2018-09-12 by Kevin Feasel

Dave Mason continues his look at lists in R:

Subsetting the list with single brackets [] for the first element returns “Atlantic”. But if we take a closer look using the str() function, we see R returned the data as a class of type list:

> #Appears to return "Atlantic" as a character class.
> division[1]
$Name
[1] "Atlantic"

> #str shows us the return is actually a list of 1 element.
> str(division[1])
List of 1
 $ Name: chr "Atlantic"

Dave also explains the difference between single brackets and double brackets for list elements.

Comments closed

Naive Bayes Against Large Data Sets

Published 2018-09-12 by Kevin Feasel

Catherine Bernadorne walks us through using Naive Bayes for sentiment analysis:

The more data that is used to train the classifier, the more accurate it will become over time. So if we continue to train it with actual results in 2017, then what it predicts in 2018 will be more accurate. Also, when Bayes gives a prediction, it will attach a probability. So it may answer the above question as follows: “Based on past data, I predict with 60% confidence that it will rain today.”

So the classifier is either in training mode or predicting mode. It is in training mode when we are teaching it. In this case, we are feeding it the outcome (the category). It is in predicting mode when we are giving it the features, but asking it what the most likely outcome will be.

My contribution is a joke that I heard last night:  a Bayesian statistician hears hooves clomping the ground.  He turns around and sees a tiger.  Therefore, he decides that it must be a zebra.  First time I’d heard that joke, and as a Bayesian zebra-spotter, I enjoyed it.

Comments closed

Triggers: Good, Bad, Mostly Ugly

Published 2018-09-12 by Kevin Feasel

Bob Pusateri walks us through a poorly-written DDL trigger:

First, the scope. While the application that deployed this trigger has its own database, AppDB, this trigger is firing for events on the entire server, which is what the ON ALL SERVER line means. Any qualifying event on this server, even if it pertains to another application with a separate database, will be written into this database. And what is a “qualifying event”? Literally any DDL statement. The line AFTER DDL_EVENTS specifies the very top of the event hierarchy used by DDL triggers.

So to recap on scope, this application is capturing all DDL statements on the entire server and saving a copy for itself. This application is seeing (and recording) plenty of events that it has no need to see. If this were a healthcare application or a system that dealt with PII it would be a legal nightmare, but fortunately it isn’t.

However, scope isn’t the only issue.

Worth the read.  If you use DDL triggers on the instance level, make sure you know what you’re looking for and limit yourself as much as possible.

Comments closed

Tracking Database Logins: 5 Methods

Published 2018-09-12 by Kevin Feasel

Eugene Meidinger has a medley of options for tracking server logins:

I once had to some auditing for a customer and it was a complicated, multi-stage process. We had to be able to demonstrate who had admin access and what kind of activity was going on, on the server. But before we could do any of that, we first had to identify who was actually logging on.

We get a brief walkthrough of each, and an important warning.

Comments closed

Server-Level Triggers

Published 2018-09-12 by Kevin Feasel

Shane O’Neill makes me wish Policy-Based Management ever got the love it needed:

Now, my normal attitude with regard to triggers tends to run to the negative. Which is horrible, because triggers are just like any other tool; neutral by themselves and only good or bad based on how we use them.

So, with that being said, I’ve forced myself to think of a positive use for them. So here is a time when I’ve used triggers for a “good” cause and used them to get some visibility on when new databases are created.

DDL triggers can be useful things, as Shane shows us.

Comments closed

Saving Table History With Triggers

Published 2018-09-12 by Kevin Feasel

Bert Wagner shows us a way of saving table history in SQL Server using triggers:

Triggers are something that I rarely use.  I don’t shy away from them because of some horrible experience I’ve had, but rather I rarely have a good need for using them.

The one exception is when I need a poor man’s temporal table.

Check it out.  My main comment is, make sure you write the triggers to handle updating multiple rows; otherwise, you’ll be disappointed when rows go missing.

Comments closed