How to scale down databases in Yandex. Cloud without downtime. An example with three hosts is

+ 0 -

Optimization of relational databases without downtime on the example of the most loaded database in Badoo

 
3r3-31. Optimization of relational databases without downtime on the example of the most loaded database in Badoo 3r33333.  
3r33333.  
Under highload conditions, the complexity of optimizing relational databases increases by an order of magnitude, since buying even more powerful hardware is expensive and it is no longer possible just to turn off the application at night for a long DB alter process and data migration. 3r33333.  
3r33333.  
We recently told how we are optimized the PHP code of our application . Now came the turn of the article about how we completely changed the internal structure of the most loaded and important database in Badoo, without losing a single query. 3r33333.  
...
+ 0 -

Oracle Bitmap Index Techniques

Oracle Bitmap Index Techniques 3r3196. 3r3-31. And again, good evening! 3r3181.  3r3196. 3r3181.  3r3196. We start the second stream of our new course 3r336. "Relational DBMS" 3r3184. which we did a little bit after the first run: additional classes on MySQL and Postgres clusters, docker was in demand, and various “file improvements”. So wait for open lessons (in which part of old topics have been carried out) and interesting materials. Today we delve into the techniques of Oracle. 3r3181.  3r3196. 3r3181.  3r3196. Go. 3r3181.  3r3196. 3r3181.  3r3196. Oracle Bitmap indices are very different from standard B-tree indices. In ...
+ 0 -

Forced optimization of PostgreSQL queries

 3r-33199. 3r3-31. What to do when there is a closed source application that does not access the database in the most optimal way? How to pull requests without changing the application, and possibly the database itself? 3r3186.  3r-33199. 3r3186.  3r-33199. If you haven't asked such questions, you are a very successful and strict DBA :)
 3r-33199. Well, if you ask, let me share my suffering and experience. 3r3186.  3r-33199.
3r3186.  3r-33199. 3r3186.  3r-33199.

Need to store more data, or setting a problem

3r3186.  3r-33199. 3r3186.  3r-33199. This section can be safely browsed if the history ...[/h]
+ 0 -

Databases and Kubernetes (review and video report)

+ 0 -

Between Security and Paranoia: Trends in Large Corporations

 
3r3-31. Monitoring the lives of large corporations drives me into a depression. This is a wild paranoia and at the same time terrible, gaping security holes. However, perhaps these things are just connected - because the paranoid is focused on certain things, and can easily overlook the obvious. He can go out into the street, frantically rustling with foil, which he has wrapped from head to foot, and be hit by a bus.
 
 
Between Security and Paranoia: Trends in Large Corporations  
 
I’ve seen a company where they closed the USB data storage profile for VDI machines, but they didn’t close the USB Hub profile, that is, you could plug in the USB Hub and then a USB ...
+ 0 -

How to run SQL Profiler Trace at night, at a specific time?

 3r3182. 3r3-31. How to start the SQL profiler trace when the problem needs to be caught from 3:00 to 3:30 in the morning? This can be done using server-side tracing, but this is extremely inconvenient. It is not difficult, but uncomfortable, and always lazy. Finally, I decided to automate this once and for all. Like this:
 3r3182.
 3r3182. How to run SQL Profiler Trace at night, at a specific time?  3r3182.
 3r3182. Jenkins here, by the way, is completely optional and serves only as an interface to invoke the script with the necessary parameters:
 3r3182.
 3r3182.  3r3182.
 3r3182...
+ 0 -

Consistency and ACID guarantees in distributed storage systems

Consistency and ACID guarantees in distributed storage systemsDistributed systems are used when there is a need for horizontal scaling to provide increased performance indicators that a vertically scaled system cannot provide for adequate money.
 
 
Like the transition from a single-threaded paradigm to a multi-threaded one, migration to a distributed system requires a kind of immersion and an understanding of how it works inside, what you need to pay attention to.
 
 
One of the problems that confronts a person who wants to migrate a project to a distributed system or start a project on it is which product to choose.
 
 
We, as a company that has “eaten ...
+ 0 -

Search for a damaged object by the number of the damaged page in MS SQL Server 2005

Search for a damaged object by the number of the damaged page in MS SQL Server 2005 3r3r1616. 3r3-31. The other day, one of the MS SQL Server databases went to Suspect, there was an error message in the log: 3r3151.  3r3r1616. 3r33939. Msg 710? Level 2? State ? Line 14
 3r3r1616. Database ID ? page (1: 386499), slot 0 for LOB data type node does not exist. This is usually caused by transactions that can be read uncommitted data on a data page. Run DBCC CHECKTABLE. 3r33939. 3r3151.  3r3r1616. The base was transferred to Emergency and attempted to perform DBCC CHECKDB, but execution stopped immediately: 3r3151.  3r3r1616. 3r33939. Msg 892? Level 1? State ? Line 13
 3r3r1616. Check ...
+ 0 -

Problems of access to personal data on behalf of all participants in the process

 3r33320. 3r3-31. Greetings dear habrovchane. 3r3308.  3r33320. 3r3308.  3r33320.
Lyrical digression [/b]
As a user of one of the “highly developed” resources of the Runet, I discovered a fairly common picture in the admin area of ​​my profile: spam was sent on my behalf. 3r3308.  3r33320. For me, this became some kind of signal, because during that time (an active user of Internet resources from 1999) while I used access methods to my resources in the form of a login and password, I did not have a single (at least registered) case of hacking or leakage of such data. For each resource I use a unique ...
+ 0 -