News & Blog back

Subscribe

Filters

PostgreSQL Streaming Replication Cheatsheet.

Print it out!

In February 2016, I will be talking about Streaming Replication at PgConf 2016 Russia.
Meanwhile, my work in progress and I want present a small cheatsheet about streaming replication. If you have some difficulties with this topic, you can print out the cheatsheet and place it on the wall.

postgresql streaming replication cheatsheet

PDF version available here.

Comments: 2

2 responses to “PostgreSQL Streaming Replication Cheatsheet.”

A.Shabanov says:

January 20, 2016 at 7:01 pm

IMO, hot_standby_feedback=on is ok when standbys working as load balancing databases. But remember your tables got bloated once standby disconnects by any reason. Vacuuming will be deferred till standby comes back. If you use SR for HA I'd not enable it by default as too much risk of bloating heavily updated tables.
Thanks.

Reply
Alexey Lesovsky says:

January 21, 2016 at 7:03 am

> Vacuuming will be deferred till standby comes back.
I don't agree with you. Where you find this info? If standby goes down, autovacuum will work as usual.
Check out my post about how hot_standby_feedback works. http://blog.postgresql-consulting.com/2015/09/postgresql-hot-standby-feedback-how-it.html

The bloat risk exists only if you run long analytics queries on replicas.

Reply

Leave a Reply Cancel reply

You may also like:

Automated index bloat management: How pg_index_watch keeps PostgreSQL indexes lean.

To follow up on my introductory blog post, here is a more extensive deep dive into ins and outs of pg_index_watch. In this post, I will take…

Please welcome Pg_index_watch – a utility for dealing with index bloat on frequently updated tables.

One of the big topics in Postgres is bloat. It is something that every DBA comes across and in fact we have a few good posts on…

No UPDATE updates

Working with databases one can’t help but wonder what happens with data stored on the disk. For example you need to know this to be able to…

VACUUM на часто обновляемых таблицах

В прошлой статье я описал общие настройки AUTOVACUUM, подходящие для обычной работы. Параметры агрессивные, но позволяют работать другим процессам в нормальном режиме. Часто обновляемые таблицы – это…

We don’t want to sound like your mother, but ~~brushing your teeth~~ having your databases checked regularly is really important for your ~~health~~ application performance.

Our two-phase Premium cover is performed by DBA with 10+ years PostgreSQL experience who live and breathe Postgres.
They will ensure that your databases have the security and high availability your application needs.

– PHASE 1 –

Audit and optimisation

This is an induction phase, during which your databases will undergo a thorough review by our DBA, who will:

analyse load tune DB configuration
review your database monitoring data
review and optimise top query
assist in monitoring tool installation
set up auto-healthchecks

During this phase our DBA will be answering any questions regarding:

architecture
database configuration
OS configuration
high availability set up and replication
backup and recovery
upgrades (major and minor)
partitioning
scaling out and sharding practices
access privileges
support during migration to PostgreSQL
consult on hardware selection
locking and concurrency issues

Training

Your developers will also have an online training session run by our experts to make sure that they can tackle key PostgreSQL issues on an industry standard level.

– PHASE 2 –

Ongoing support

Once your databases have been audited and your team trained you will be moved to a subscription-based support. During this phase our DBA will be around to answer any questions and support in the following:

Questions related to database and OS configuration
Basic query optimisation
Questions that do not require an in-depth investigation
Post factum analysis of database outages (during working hours)
Autohealthcheck reports review (we notify you of any abnormalities)
Monitoring trigger review (we notify you of any abnormalities)

Price: from 11000€/year (for 2 PostgreSQL instances), Premium PostgreSQL cover is a one-year contract.

Price includes:

up to 10 hours DBA time monthly for the duration of Phase 1 (2 months) and
up to 2 hours monthly for Phase 2 (10 months).

Overtime is billed at the end of each month at 250€/hour rate.

Subject to client’s installation specifications and databases access Phase 1 may take longer and require additional fee that will be pre-agreed with client in advance.

SLA: Response time is provided as follows:

During Audit and optimisation (Phase 1) Data Egret provides support with up to 3h response during working hours.
During the Ongoing support (Phase 2) we answer any questions via email with up to 24 hours response time.

If you are interested in shorter response times we suggest our 24/7 PostgreSQL Support services.

Database Access: Data Egret PostgreSQL cover can be provided with or without access to databases, however without database access addressing any issues requires more time and therefore is less cost-effective process.

Interested in learning more? Give us a nudge via email or click the button below.

We are a consulting firm that offers remote PostgreSQL administration services. We are looking for a seasoned PostgreSQL DBA specialist for a remote full-time position.

Who will you be working with?

Some clients have their own DBA teams, with whom you will be working and some don’t – in this case you will be working with client’s developers and technical managers. In either case you will be surrounded by best in class Data Egret PostgreSQL DBA, who have many years of experience servicing our clients.

What skills will you gain?

As well as harnessing your technical skills you will grow as a technical support specialist, who is able to listen to our clients, ask questions, review possible solutions with them, and be a reliable partner, who will add value beyond customer expectations.

What will your job be?

You will be the focal point to our clients when it comes to databases – figuring out why things don’t work the way they should, providing guidance and fixes to ensure that our client’s high availability projects go like clockwork.

Here are a few examples of your day-to-day technical tasks:

Make changes in database’s schema under load
Perform slow query analysis and optimisation
Correctly create/adjust indexes for query optimisation
Set up and configure Postgres on new servers
Setup Postgres database replication
Perform failover/switchover
Set up backup process and perform recovery
Perform minor and major upgrades
Architectural consulting
Database health check

What do we expect from you?

You are a seasoned DBA with 2 or more years of experience in administering high availability PostgreSQL projects (including replication configuration, performance tuning, monitoring configuration etc.)
Have experience in optimising complex SQL queries
Enjoy working independently on vague client briefs
You are a confident Linux user and know how to administer it (Ubuntu/Debian/RedHat),
Have a general understanding of database theory
Have basic knowledge of shell scripting (writing a couple of screens-lengthy script that will be understandable to other admins and yourself a few month down the line)
You can write a complex (and very complex) SQL query
You are a keen learner and eager to discover new solutions to old problems

This is a client facing role and the ability to communicate your actions and be able to reason with clients is imperative.

What skills would be a plus?

Familiarity with Patroni
Familiarity with some DevOps tools such as Ansible/Docker/Kubernetes
C programming skills and experience in analysing with perf/strace/gdb
Administration experience with cloud platforms
Basic knowledge of Russian language would be a plus

Why choose Data Egret as your place of work?

You will become part of one of the leading PostgreSQL DBA teams working in a company that has been named a significant sponsor to PostgreSQL by the Postgres community.
Will be able to brush up your database administration skills on a multitude of databases in different environments and gain experience that would take you years to get working in-house.

What do we offer?

A competitive pay based on your experience and annual performance-based bonus scheme
Full-remote working, offering you the opportunity to actually have a life 🙂
Flexible working hours
An opportunity to work alongside colleagues DBA, who live and breathe Postgres and contribute to its development
Support of your participation in community conferences
Collaborative and open environment where you are supported by your colleagues and can learn and thrive

Think you qualify?

Have questions about this opportunity?
Contact us! [email protected]

Health Check

Most emergencies leading to service unavailability and loss of your data can be mitigated and detected at an early stage with minimal effort. Data Egret Health Check will ensure a smooth running of your service, give you a headstart for your launch or migration and will allow better budgetary planning.

We will provide you a tailored offering based on your system specifications, however, below rates will give you an idea of pricing for our Health Check solution.

Package	Price*
PostgreSQL configuration analysis	from 500 €
Hardware and OS configuration analysis	from 600 €
Slow SQL queries analysis	from 500 €
Comprehensive performance Health Check*	from 3500 €
* For single PostgreSQL installation (1 master + 1 replica)

24/7 PostgreSQL Support

24/7 PostgreSQL Support offers you a combination of ongoing consultancy from our experts as well as round the clock support from dedicated team. To ensure smooth running of your systems and give you peace of mind we will be providing following services:

Dedicated Data Egret team

that will be your contact in case of any questions or concerns. Team that will be aware of your system’s specifications and will work on any arising issues within 15 minutes to an hour from enquiry.

Real time counsel

on everything PostgreSQL provided by our team of experts to your company developers and operations engineers.

Ongoing monitoring of your databases

that includes query performance, issues management and recommendation on performance improvements.

Regular upgrades

of PostgreSQL and related software.

Backup and recovery

that includes regular backup of your database and recovery solution with regular test recovery.

Basic	Premium	VIP
Incl. up to 5h/month DBA time	Incl. up to 15h/month DBA time	Incl. up to 30h/month DBA time
Slack, Mattermost, Telegram	Slack, Mattermost, Telegram	Slack, Mattermost, Telegram
Response time — 1 Hour	Response time — 1 Hour	Response time — 30 minutes
Up to 2 PostgreSQL instances	Up to 5 PostgreSQL instances	Up to 10 PostgreSQL instances