Uber Engineering

Uber’s mission is transportation as reliable as running water, everywhere, for everyone. Last time, we talked about the foundation that powers Uber Engineering. Now, we’ll explore the parts of the stack that face riders and drivers, starting …

The post The Uber Engineering Tech Stack, Part II: The Edge and Beyond appeared first on Uber Engineering Blog.

Overview

When working with Amazon AWS Aurora, there are some steps to consider when trying to get data out of an active Aurora master into a slave, potentially into a EC2 instance or offsite in another data centre. Creating an external mysql to Aurora gives the option to move out of Aurora, or to have the flexibility to move data around as desired. With AWS RDS instances this task is pretty simple because you can do the following :

1. Create a read replica
2. Stop the slave process
3. Capture the positioning
4. Dump the database

With Aurora it’s a little trickier, because a read replica in Aurora has no slave process. All of the replication is handled on the back end and cannot be controlled. However, setting up an external slave can be done.

Amazon AWS Documentation

In …

When it comes to building database applications and solutions, developers, DBAs, engineers and architects have a lot of new and exciting tools and technologies to play with, especially with the Hadoop and NoSQL environments growing so rapidly.

While it’s easy to geek out about these cool and revolutionary new technologies, at some point in the development cycle you’ll need to stop to consider the real-world business implications of the application you’re proposing. After all, you’re bound to face some tough questions, like:

Why did you choose that particular database for our mission-critical application? Can your team provide 24/7 support for the app? Do you have a plan to train people on this new technology? Do we have the right hardware infrastructure to support the app’s deployment? How are you going to ensure there won’t be any bugs or security vulnerabilities?

If you don’t have a plan for …

MySQL-Sandbox 3.1.11 introduces a new utility, different from anything I have put before in the MySQL Sandbox toolkit.

make_sandbox_from_url downloads a tiny MySQL tarball from a repository and install it straight away.

As of today, the following packages are available

In this blog, I intend to present the MySQL Enterprise Monitor (MEM), and also to present a multi-machine vagrant setup by adding a vagrant-hostmanager plugin to configure the /etc/hosts on the VMs. To start, I begin by installing the hostmanager plugin on my mac where I manage my environments. [crayon-58c2f17573ee8441268854/] OK, so that was easy.… Read More »

Awhile ago I found myself analyzing a MySQL fabric installation to understand why a group member was occasionally being marked as FAULTY even when the server was up and running and no failures were observed.  

         
                         server_uuid     address  status       mode weight
------------------------------------ ----------- ------- ---------- ------
ab0b0653-6121-11c5-55a0-007543445454 mysql1:3306 PRIMARY READ_WRITE    1.0
f34dd331-2432-11f4-a2d3-006754678533 mysql2:3306 FAULTY  READ_ONLY     1.0

Upon reviewing mysqlfabric logs, I found the following warnings were being logged from time to time:

[WARNING] 1442221217.920115 - FailureDetector(xc_grp_1) - Server (f34dd331-2432-11f4-a2d3-006754678533) in group (xc_grp_1) is unreachable

Since I was not clear under which circumstances a server is marked as FAULTY, I decided to review MySQL Fabric code (Python) to better …

Replication between Tungsten clusters

The process I will describe in this post will allow you to configure replication between Tungsten clusters. The most common use case I have seen for this is a dedicated ETL cluster.

The setup will look like this:

So we will have a composite datasource (compositeprod) composed of east and west clusters, and two ETL clusters, one on each side.

The described setup allows failover or switchover within a single datacenter (e.g. db1.east -> db2.east) or to the Disaster Recovery Site, (e.g. db1.east -> db1.west) for the core cluster.

At the time of this article’s publication it is not possible to replicate between two composite clusters, so each ETL cluster needs to be standalone. This feature is expected for Tungsten 5.0 release.

We will install a standalone replicator on the ETL hosts to bring data in …

Overview

Cloud storage is becoming more and more popular for offsite storage and DR solutions for many businesses. This post will help with those people that want to perform this process for MySQL backups directly into Amazon S3 Storage. These steps can probably also be adapted for other processes that may not be MySQL oriented.

Steps

In order to perform this task we need to be able to stream the data, encrypt it, and then upload it to S3. There are a number of ways to do each step and I will try and dive into multiple examples so that way you can mix and match the solution to your desired results.  The AWS S3 CLI tools that I will be using to do the upload also allows encryption but to try and get these steps open for customization, I am going to do the encryption in the stream.

1. Stream MySQL backup
2. Encrypt the stream
3. Upload the stream to AWS S3

Step 1 : …

Since documentation is not very clear to me on the topic of InnoDB flushing in combination with Linux IO (specifically the write system call), I decided to put together this article in hopes of shedding some light on the matter.

How Linux does I/O

By default, the write() system call returns after all data has been copied from the user space file descriptor into the kernel space buffers. There is no guarantee that data has actually reached the physical storage.

The fsync() call is our friend here. This will block and return only after the data and metadata (e.g. file size, last update time) is completely transferred to the actual physical storage.

There is also fdatasync() which only guarantees the data …

It is not uncommon to come across MySQL databases where reserved words are in use as identifiers for any kind of database objects.

Perhaps when the application schema was implemented, the words were not reserved yet, and they became reserved later on a subsequent MySQL release.

It is a good practice to check reserved words usage prior to doing any database upgrades, as any newly reserved keywords will cause syntax errors on the new version.

This is usually not a problem if proper quoting is used for referencing the objects, as described on the official manual page.

The actual steps to do this depend on the environment; for example, the following can be configured to tell Hibernate to escape identifiers:

property name="hibernate.globally_quoted_identifiers" value="true"

This does …