Working With Lists In Python

Jean-Francois Puget shows that iterating through lists in a for loop is not always the most efficient for Python code:

The above code is correct, but it is inefficient.  List comprehensions are meant for case like this.  The idea of comprehension is to move the for loop inside the construction of the result list:

def get_square_list(x): return [xi*xi for xi in x]

This is both simpler and faster than the previous code.

Definitely worth reading if you come at Python from a C-like language background.

Caching Strategy

Kevin Gessner explains some caching concepts used at Etsy:

A major drawback of modulo hashing is that the size of the cache pool needs to be stable over time.  Changing the size of the cache pool will cause most cache keys to hash to a new server.  Even though the values are still in the cache, if the key is distributed to a different server, the lookup will be a miss.  That makes changing the size of the cache pool—to make it larger or for maintenance—an expensive and inefficient operation, as performance will suffer under tons of spurious cache misses.

For instance, if you have a pool of 4 hosts, a key that hashes to 500 will be stored on pool member 500 % 4 == 0, while a key that hashes to 1299 will be stored on pool member 1299 % 4 == 3.  If you grow your cache by adding a fifth host, the cache pool calculated for each key may change. The key that hashed to 500 will still be found on pool member 500 % 5 == 0, but the key that hashed to 1299 be on pool member 1299 % 5 == 4. Until the new pool member is warmed up, your cache hit rate will suffer, as the cache data will suddenly be on the ‘wrong’ host. In some cases, pool changes can cause more than half of your cached data to be assigned to a different host, slashing the efficiency of the cache temporarily. In the case of going from 4 to 5 hosts, only 20% of cache keys will be on the same host as before!

It’s interesting reading.

Network And Sankey Diagrams In Python And R

Tony Hirst has a roundup of various R and Python packages which build network charts or Sankey diagrams:

Another way we might be able to look at the data “out of time” to show flow between modules is to use a Sankey diagram that allows for the possibility of feedback loops.

The Python sankeyview package (described in Hybrid Sankey diagrams: Visual analysis of multidimensional data for understanding resource use looks like it could be useful here, if I can work out how to do the set-up correctly!

Sankey diagrams are on my list of dangerous visuals:  done right, they are informative, but it’s easy to try to put too much into the diagram and thereby confuse everybody.

Predicting Restaurant Reservations With A Neural Net

Kevin Jacobs builds a simple neural net using Pandas and sklearn:

The first thing to notice is that our values are not normalized. The number of visitors is a number and gets larger and larger. To normalize it, we simply divide it by 100, since all numbers are below 1. The same holds for the lag. Most of the lags are lower than 30. Therefore, I will divide the lag size by 30.

Notice that there are many more approaches for normalizing the data! This is just a quick normalization on the data, but feel free to use your own normalization method. My normalization process is closely related to the MinMaxScalar normalization which can be found in sklearn (scikit-learn).

With just a few lines of Python code we can create a Multi-Layer Perceptron (MLP):

Click through for the code.

“Caveman” Graphs In SQL

Denis Gobo puts together some basic Management Studio data visualization:

I found this technique on Rich Benner’s SQL Server Blog: Visualising the Marvel Cinematic Universe in T-SQL and decided to play around with it after someone asked me to give him the sizes of all databases on a development instance of SQL Server

The way it works is that you take the size of the database and then divide that number against the total size of all databases. You then use the replicate function with the | (pipe) character to generate the ‘graph’  so 8% will look like this ||||||||

You can use this for tables with most rows, a count per state etc etc. By looking at the output the graph column adds a nice visual effect to it IMHO

It does the job and doesn’t require you to go out to a different product, so it works pretty well for occasional administrative queries.

Locks And Partitioning

Erik Darling looks at the confusing mess that is SQL Server partitioning:

In the Chicago perf class last month, we had a student ask if partition level locks would ever escalate to a table level lock. I wrote up a demo and everything, but we ran out of time before I could go over it.

Not that I’m complaining — partitioning, and especially partition level locking, can be pretty confusing to look at.

If you really wanna learn about it, you should talk to Kendra — after all, this post is where I usually send folks who don’t believe me about the performance stuff.

Click through for that demo and explanation.

Getting Started With Always Encrypted

Monica Rathbun kicks off a series on Always Encrypted:

There are two possibilities Deterministic and Randomized.

MSDN defines Deterministic encryption as always generates the same encrypted value for any given plain text value. Which means that if you have a birthdate of 01/03/1958 it will always be encrypted with the same value each time such as ABCACBACB. This allows you to index it, use it in WHERE clauses, GROUP BY and JOINS.

Randomized encryption per MSDN- uses a method that encrypts data in a less predictable manner. This makes Randomized encryption more secure, because using the example above each encrypted value of 01/03/1958 will be different. It could be ABCACBACB, BBBCCAA, or CCCAAABBB. All three encrypted values are subsequently decrypted to the same value. Since the encrypted value is random you cannot perform search operations etc. as you can with Deterministic.

Part 1 is about building the certificates and keys needed to encrypt data.

Upgrading That Expired Evaluation Copy Of SQL Server

Cody Konior finds a way to extricate the poor souls who need to upgrade expired evaluation copies of SQL Server from their mess:

Common advice here is to set the clock backwards. My problem with that is that you’re probably doing this on an unsupported unknown black-box flaming garbage can of a system set up by someone who wasn’t meant to do it – because otherwise they wouldn’t be using the evaluation edition. So what are the repercussions of setting the clock backwards? Perhaps their application spawning silently in the background and trashing this or other databases with bad date information? Perhaps you’ll lose your RDP connection and then be unable to connect back in because of the SSPI error generated by a clock mismatch?

No thanks. Instead you need to do some detective work.

Read the whole thing.

Pre-Attentive Attributes

Meagan Longoria continues her discussion of design concepts with pre-attentive attributes:

Four preattentive visual properties have been defined:

  • Color (intensity, hue)

  • Form (orientation, line length, line width, size, shape, curvature, enclosure, added marks)

  • Spatial Positioning (2-D position)

  • Movement

Good information, and then Meagan ties it to Power BI.

Using Pester For Configuration Checks

Andrew Pruski shows how to use Pester to audit SQL Server configuration settings:

One Pester test running!

What I like about this is that it can be easily dropped into a job scheduler (e.g.- Jenkins) and then you’ve got a way to routinely check (and correct) all the configuration settings of the SQL instances that you monitor.

Pester would not have been my first thought for configuration checking, but it does serve as another useful option.

Categories

December 2017
MTWTFSS
« Nov  
 123
45678910
11121314151617
18192021222324
25262728293031