Universal jam

Heaps and bleeps

I really has been a while since I have posted something here.

But now, after figuring out some administrative details, is time for another post.

How do we structure our data? How is it organized?

The question of structuring data and especially organizing it in the most efficient manner on storage is a century old question.

Therefore, indexes (whether they represent the table data itself or they represent duplicated subsets of a table) are here to help us organize the data for a better accessibility and better performance.

The topic of heaps has been chewed on for at least a century, as well. No doubt about that.

But somehow I keep thinking that there are plenty of heap tables left out there, especially in some legacy DB systems, produced by big third party vendors (wink, wink), which were built at least some 10 years ago, which were not even imagined to work with millions of rows.

Anyhow, the following script will give you a result set of all heap tables and their row count and usage.

I.e. you should hear a bleep (well, make sure your speakers are on!) if this script returns some tables with row count more than several thousand.

If it does, then go to your DBA, or to your third party vendor, blow a whistle and ask WHY.

Here is the script, keep in mind that it is accessing the metadata of one database at a time, hence you have to use it in context.

 

 

?Download download.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22

SELECT  SCHEMA_NAME(o.schema_id) AS [schema] ,
        OBJECT_NAME(i.object_id) AS [table] ,
        p.rows ,
        user_seeks ,
        user_scans ,
        user_lookups ,
        user_updates ,
        last_user_seek ,
        last_user_scan ,
        last_user_lookup
FROM    sys.indexes i
        INNER JOIN sys.objects o ON i.object_id = o.object_id
        INNER JOIN sys.partitions p ON i.object_id = p.object_id
                                       AND i.index_id = p.index_id
        LEFT OUTER JOIN sys.dm_db_index_usage_stats ius ON i.object_id = ius.object_id
                                                           AND i.index_id = ius.index_id
WHERE   i.type_desc = 'HEAP'
        AND ( user_seeks IS NOT NULL
              OR user_scans IS NOT NULL
              OR user_lookups IS NOT NULL
            )
ORDER BY rows DESC

 

 

January 27th, 2012 | Category: Performance audit, SQL Server metadata, T-SQL | Leave a comment
Work on itWell, now...OKGoodGood job! (1 votes, average: 5.00 out of 5)

If you gave a rating less than the maximum, you must leave a comment with suggestions on how to improve the post. :)

Loading ... Loading ...

 

Top five reasons why good employees leave their jobs

There are many different cultures around the world and they are closely related to the work ethics and the employer-employee relations.

A while back I wrote an article on Interviewing (how to gather a team of great employees). But gathering a great team is not enough – the team has to be maintained, developed and kept together.

Today I will write about the five greatest turnoffs from the employee’s point of view, i.e. why do employees leave their jobs.

 

As I mentioned earlier, the habits are closely dependent on the surrounding environment and on the culture of the country. For example, it is quite well known that in Sweden people are less prone to leave their jobs and to look for other opportunities compared to south-European countries. Recent surveys show that some 30% of the employees in Sweden were unhappy with their jobs, but even then people do not tend to look for other job. There was even a case where a person was treated really poorly by their colleagues and did not leave the company for 10 years. (Job security aside – but you must agree with me that there is no reason to be treated poorly for 10 years without doing anything about it.)

Anyhow, regardless of the cultural differences, I think that the following 5 reasons are quite standard and adequate in most – if not all – cultures. The points below are not sorted in a particular order, and certain aspects of the employment are omitted – for example: being underpaid, not receiving help with study materials etc. I know that there are a lot of underpaid and unfairly treated professionals out there, but at the end it is all about moving foreward.

If a company does not want to move forward by improving the environment for their employees, then it gets harder to keep the employees anyway. So here is the list:

 

  1. Bureaucracy – in my years of experience I have noticed that only bureaucrats like bureaucracy; IT professionals do not like it. Employees generally do not like to repeat the same information to several bosses or managers, and do not like to be talked to by several managers about the same thing, either. The simple reason is, that sooner or later the situation turns into that famous “red stapler” scene from “Office Space”.
  2. Poor / flawed communication manners – a manager / company should not wait for their employees to “send signals”; even worse – a compnay should not “capture” signals from the employees. The best way is to have one-on-one meetings with the employees occasionally and talk to them. The best way to do it is to invite an employee to lunch from time to time and to talk to them in a friendly unengaging atmosphere. Having bulk-meetings will never show accurate readings due to the very nature of the human being: people behave differently socially and alone.
  3. Suppressing creativity – never tell your employees that they should not take initiative or that they should not express their tallent in any way. Mistrusting the professional urges and the pure intentions of the employees will only put the company in a very negative light which eventually will eat up and weaken all links.
  4. Taking too much of the personal time of employees – it is a proven fact that it is not healthy (and actually almost impossible) for a human being to concentrate and be involved in work related activities for 8 hours without any break. If an employer / manager takes too much of the employees’ attention, or if they are deprived of clean and quiet private moments during the work day, then the employees get quite sour and unproductive.
  5. Not following up on the employees’ professional improvement – a company benefits greatly from the employees’ professional improvement, but the more advanced the employees get professionally, the harder it is to keep them around. It is true – there is always some other company that pays better, or has extra vacation days or some other benefits that attract employees. Keeping employees restricted and averagely improving should not be used as a way to retain employees. Instead, giving feedback and making employees feel special is the best way to keep them around.

Finally, keep in mind that the employees’ productivity is proportional to what their perception of the importance of their job is. This is the top factor which defines productivity long term. Money, perks and gadgets come after that…

 

 

 

 

January 11th, 2012 | Category: Ramblings | Leave a comment
Work on itWell, now...OKGoodGood job! (No Ratings Yet)

If you give a rating less than the maximum, you must leave a comment with suggestions on how to improve the post. :)

Loading ... Loading ...

 

A few words on social media and why it is evil

I usually try to post only technical articles on this blog, especially articles which are related to SQL Server and databases. This one is a bit more general, even though the topic of data warehousing and analysis comes up a few times.

Lately I have been thinking about the social media and its role in our social context. Can’t really help it – I have to share a few thoughts.

Let me first mention that this blog post is not necessarily about Facebook, nor Google+, nor LinkedIn in particular. By ’social media’ I mean ‘any public medium of computerized source where the user may choose to spill personal information, feelings and intentions’.

 

What is wrong with that. Well, a few things.
Lack of personal choice – it is true that creating a profile or account in the social media is voluntary (aside from the peer pressure, of course). It is also voluntary to post updates, picures, opinions and intentions.

From the point of view of a database specialist, let me make a comment here:

The rule number one in database warehousing is: never ever ever delete any data.

Having said that, we come to the point where it is voluntary to share your urges, but once they are shared, they are not yours anymore. Once shared, they are owned forever by the database owner (remember the old warning: only the written words remain forever).

And since databases are generally written to disks, the disks also retain the data even if the data is deleted later on. (I will not go in details here, but keep in mind that deleting a file from a hard disk does not always mean that the file is lost – there are many ways to recover it anyway.)

So, no matter what you do, as soon as you click the ‘Post’ button, you are stuck forever with admitting (and committing to) your statements.

Here is another comment from the point of view of a database specialist:

The rule number 2 of the database warehousing is: since you never delete any data, then you’d better do something with it!

So,  what’s there to do with all this data?

Trend it. Trend it good. Sell it. To whomever seems interested.

And if you don’t have enough buyers of the trended data – make sure to collect some more data, until the buyers become interested enough.

And here is how it goes: you post your opinions, likes, dislikes, your location, gender, line of work, the list of your friends and your social circles. Then the data is collected and trended, some of your browsing history is also collected and trended. The trended data is fed to a powerful rich uncle, which works most likely in the marketing sphere. As a result, the next time you visit your favourite social media you get targeted by advertisement. The data of this advertisement targeting is also collected (how long you stayed on the page, did you hover over the ad, did you click on it). The trending yet again is fed to the rich uncle who will refine the advertisement targeting next time.

So, it turns out that you kinda work for the rich uncle without being paid.

 

Another aspect of the social media is its layout in our daily timelines.

I.e.: it sucks up your time – it is a strange feeling, but sometimes people get more involved with posting their statuses than with their daily work.

Let’s suppose that you spend 1 hour per day actively using the social media sites. This is just about 6% of your waking hours.

Where does this time come from? Where does it go?

Does it come from your work time? Does it come from your sleep? (I.e. do you sleep less or work less to compensate for the 6% social media overhead?) Or do you account it as a relaxation?

And it really gets scary when you scale it: 1,000,000 users of the social media spend 1,000,000 active productive man hours on posting their statuses, opinions and intentions. PER DAY. (And this is in a best case scenario – I personally know people who spend much more than that.)

And the funny thing is – I don’t think that the romans would have built much if they had social media at the time.

And the Ottoman empire would not have destroyed as much – oops, I mean – acquired as much territory if there was social media data to be trended and analyzed by the surrounding lands.

And so on…

Getting back to the topic of trending and analysis of the social media data.

Social opinions trending – not necessarily correct, but almost there.

Recently it has shown that it is almost impossible to watch CNN or any news channel, in that matter, which does not quote data analysis from the social media.

For example, when there is a social disturbance, the social media immediately trends and reports the general opinion and the estimated participation level.

So, I guess, the social media makes it easier to handle – if you know what the social opinion is and if you know how many participants are sharing the opinion, then actions can be taken before the problems escalate, or maybe even some ideas can be prevented from propagation.

Wait a second: wasn’t this somehow mentioned in 1984? Or am I dreaming?

On the bright side, the social media is not accurate (and it never will be) in measuring the social opinions, trends and intentions, since it measures only the opinions of the subscribers, and not of the society.

Unfortunately, the above statement is not openly discussed and never put in a disclaimer when the social media trending is quoted.

Furthermore, the social media provides APIs, which give users the opportunity to create applications and to perform actions on behalf of the members of the social media. In other words, an application may actually post a status message (an opinion or even an intention) on behalf of the users and then the trending may be skewed.

 

So, let’s recap: the main activity of mankind nowadays is to create a vast collection of trended information, opinions and intentions, thus creating (almost) predictible societies.

Why not!

After all, the world problems which are not solved yet are not solved because noone really wanted to solve them too much.

Think about it: people wanted to go to Mars and by wiorking hard it was accomplished.

The poverty, starvation and lack of education, on the other hand, in big parts of our planet, though, have remained unresolved and poorly addressed.

It’s all about priorities.

Think about it.

 

 

January 9th, 2012 | Category: Data Generation, Data Warehousing, Ramblings | 4 comments
Work on itWell, now...OKGoodGood job! (4 votes, average: 4.00 out of 5)

If you gave a rating less than the maximum, you must leave a comment with suggestions on how to improve the post. :)

Loading ... Loading ...

 

 
  Older Entries »