Monthly Archives: July 2015

One year as a “Head of Data”

This happened a year ago – I’ve became a Technical Lead for the database development in the SE department. It has been a memorable year, indeed…

Last week I’ve presented a talk about this one year – what I was planning to do, what I’ve done, and what I am going to do in the next year and in the next two years. While I was preparing this talk, I’ve talked to all the squad leads, asking them, what I am doing right, what I am doing wrong, what they want me to do more, and what they want me to do less… I’ve also reviewed my first presentation in my new role, which I’ve delivered in the beginning of August last year.

You know, what was the most surprising? Every day through all this year I had a feeling, that I am not doing enough, that just like Alice in the Wonderland, I am running and running just for the sake of staying on the same place, and sometimes really falling behind. But when I looked at this my last year presentation, when I’ve started to go item by item on what was done, I’ve said to myself: wow! I actually did something!

I feel that my most important accomplishment over this year is the fact, that everybody just forgot, what was the most pressing problem a year ago. I’ve asked a number of my coworkers: do you remember, what was the most important problem for me a year ago? Do you remember, what everybody expected me to resolve? And nobody can recall!

And this is really good – because the most pressing problem a year ago was to “teach” the database developers and application developers to work together, to be collaborative, do not use the terms “us” and “them” (I really hate when people say this!) And yes, these days the database developer had blended into the software development squads so well, that nobody remember there was ever a problem.

There are many other things we’ve accomplished over the past year; as many of the lead application developers mentioned to me “now we are much better in dealing with the data, we know what to do, and what not to do”. The quality of SO development increased drastically, and the application developers know when to call for db dev advice, and are not afraid of this “whole bunch of SQL code”. We increased the amount and quality of testing. We made significant progress in fighting the Object-relational Impedance Mismatch.

And as usual – still a long way to go! But now I am pretty confident – we can do it. We have ton’s of cool stuff in works, and we have amazing people. As I’ve already mentioned a couple of weeks ago, I am going to showcase so of the work completed by my coworkers (since they are not doing it themselves :)). Stay tuned!

Leave a comment

Filed under Companies, People

PG Open 2015

This time the PG Open won’t be in Chicago, which makes me sad – I love to show-off our beautiful city. But good news – first time since 2000 I will have an reason to be in Dallas TX!

So for those who are interested – here is the conference web site, and I will be there!

Leave a comment

Filed under events

What I am busy with these days

I haven’t been posting anything for a while. I am still crazily busy at work, yet I’ve being feeling, that what I am doing at work right now is of almost no interest to anybody. And this feels paradoxic, because I have no doubt, that the work I am doing now is really important for company’s success. So I decides to try to break my silence and write a couple of short posts describing what I’ve being doing, hopefully it won’t be boring!

About three weeks ago we completed a very important project. I am saying “we”, although I personally did nothing, I didn’t write a single line of code. Yet all the people involved told me, that this project won’t happen without me.

The scope of the project was to clean up the objects ownership and permissions on all the database clusters which run our new platform. Doesn’t sound exciting, right? Could we continue operating without this cleanup? Definitely. But these discrepancies made many production deployments unnecessary difficult, because each time “something” might not work on “some” clusters. Besides, we could not set up the sating environment to look the same as production permission-wise.

This project called for the joint effort of DBAa, database developers and lead application developers. We did test the process the best possible way, but since we didn’t have an exact match of staging and production environment, there was still a possibility that something unexpected would happen. We’ve being switching ownerships and permissions one cluster at a time, with the system owners closely monitoring the execution of the apps. It took us three sessions to complete all the changes, and for most of the developers and the business users nothing really happened. Nobody noticed anything (which is a good thing, but then you feel like nobody really appreciate what you’ve accomplished:)).

I’ve got a moment of a real satisfaction when a couple of days later a very complex permission-sensitive deployment (one of those which would always cause pain for the application owners) went through without any issues!

Now we can model our staging environment from the correctly set production environment, and make all future permission changes even more safe and secure.


Filed under Data management, Systems