Мы используем куки и предполагаем, что вы согласны, если продолжите пользоваться сайтом
Понимаю и согласен
Заявка на обратный звонок
Свяжемся с вами в течение рабочего дня
Нажимая кнопку, вы соглашаетесь на обработку ваших персональных данных согласно нашей политике конфиденциальности.
We administer
highload servers
We major in DevOps, CI/CD, Kubernetes. We configure and support servers with web, backend, data bases and telephony.
Мы используем куки, чтобы предложить персонализированное содержимое и анализировать наш трафик. Прочитайте, как мы используем куки и как вы можете их контролировать через "Настройки куки". Мы предполагаем, что вы согласны, если продолжите пользоваться сайтом.
We deliver such services as:
DevOps, CI/CD, Kubernetes. We will support even single server.
For classical administration
Complex administration
If a project includes CI/CD, IaC, configurations management, clusters, fault-tolerance, then it demands highly qualified specialists to support it.

The extra charge for additional complexity can vary between
20 000 and 100 000 Rub/month
We name the exact price when we see the full list of issues.

Our production standarts
We prefer to do initial setup
A server configured by someone else can be a mine-field for us. There can be found unpredictable software versions, or unmanageble configurations, or well-disguised backdoors. We are responsible for your servers security and availability and, thus, we install our own OS and SW, and configure your servers by our own standarts.

If a client doesn't want us to do an initial setup, we then support the project as is, but in the contract there would be a clause saying what we can not be hold responsible for.

You can switch to our stantard support level anytime and accept terms with our initial setup.
We have our standart set of software
We prefer CentOS.
We use systemd-nspawn and openvz for simple web projects due to low overhead, simple and fast hardware migration and fast backup recovery.
We build CI/CD projects using k8s, Docker, Gitlab.
We use ansible for configuration management.
We are highly experienced in working with most popular DBs, such as as MySQL (Percona Server, MariaDB), and PostgreSQL, and NoSQL (Redis, MongoDB). We build HA and FT solutions using DB clusters.

We chose Gitlab as a CI/CD tool, but we can use Jenkins and TeamCity too.

For analytics and monitoring, we use Zabbix and Prometheus. But we can use M/Monit, Elasticsearch, Fluentbit/Fluentd too.

For developing of automatisation scripts, we chose Ansible.
Southbridge is a teal organisation
Our management system is horisontally oriented: Our clients deal with teams of adminitrators directly. A team is interested in understanding client's needs as well as possible, and to find best solutions possible. Thus, free from red-tape burden, our employees unlock their potential. They are also responsible for the outcome of their work to clients.

If there are any problems, our custom manager helps to reach a compromise.
Southbridge helps in developing and updating administration
We translate and write posts of our own authorship for our blog on «Habr».
We develop our own open source administration tools and put them online on our own Github.

We have developed «Slurm» workshops on Kubernetes and DevOps for system administrators.

We are CNCF Kuberneters service and training providers. As of July'19, in Russia, there are only 3 Kubernetes Certified Service Providers and 1 Kubernetes Training Provider.
An SLA can be drawn up on demand
Before signing an SLA which provides specific guarantee, a prework must be done:

Southbridge audits a project to evaluate its current state, and then to offer an infrustructure (a provider, servers quantity and characteristics) and an architecture to fit an avaliability level needed.

A client provides a suitable infrustructure and, if required, updates the project so it fits the architecture.

Done with the prework, we can sign an SLA, for we would be able to ensure your Availability:
95%, without our initial setup.
99%, with the initial setup.
99,9%, for fault-tolerant/cluster solutions (Kubernetes and others).

We can start prework before signing an SLA at any moment after a contract is signed.
SLA doesn't cover project downtime due to client's actions or omissions.
Our clients
Still have doubts?
I'm afraid to grant access to sensitive information for your administrators
We sign an NDA.
Every Southbridge employee signs the non-disclosure agreement (NDA) that imposes a heavy penalty for disclosure of sensitive information, whether it is server settings or data stored on it. Violations never happened.

I'm afraid of being hacked through you
We use several additional security enhancenments such as one time passwords (OTP, 2FA), bruteforce checks, we setup firewall in offensive mode (all that is not allowed is prohibited).
I'm afraid that an emergency will happen at night, and your administrator will postpone the problem until morning
Our duty squad will escalate any issue, which they can't do, to your administrator. We will fix the problem, even if the emergency happens on New Year's Eve.

My server is a mess. I'm afraid, you will not cope with it
Putting projects in order is what we usually do. If you want to systemise and optimise your server infrustructure and if you are willing to assist us in doing so, we will do it. But we sure are not willing to support a zoo as is.
I'm afraid, you will set up a vendor lock-in, and no one will be able to support my servers after you leave
We, in Southbridge, tend to unification and standartisation of solutions. We make a structure transparent, and all documentation is stored in our Wiki. Initially, Kubernetes was made so an administrator could easily inherite a project from those who did the job before. There is only one "but": An engeneer who is going to work when we leave has to be qualified enough to support a project of such a coplexity level.
I'm afraid that if you won't do it, I will suffer losses
If a project tends to be risky, say, if we don't know your stack, or if we lack some specific experience, we point it out straightforwardly during negotiations, so that you could decide for yourself if you are ready to try it with us or to hire another contractor.
I'm not sure if your specialists are skilled enough
We have a rich portfolio: 10 + successful CI/CD system integrations, support for 600+ dedicated servers and 30 + Kubernetes clusters. We have 16 certified Kubernetes administrators (CKA) and 10 certified Bitrix administrators. We are now Kubernetes Certified Service Providers. We were handed over to support the servers of Modulbank, REN-TV, Gold585, Azbuka Vkusa.
How are you going to ensure an uptime of 99.95%?
Using devops tools like Ansible, Docker, Kubernetes, and building CI/CD processes.
I'm afraid, my project will not have enough attention
We unify and replicate solutions. Each our client engages using proactive hardware monitoring software tool, we elliminate software vulnerabilites for all. We introduce the new technology to everyone, who is ready to use it. 164 clients gave us one server for support each, and no one's still complained about the lack of attention.
Sometimes, there can be some misunderstanding, when our administrators are waiting for a client's reaction, while a client is waiting for administrators to do something. If you think that your project lacks attention, please, send us a note on ask@southbridge.io
I'm afraid, you will deliver support for my server only externally, while a problem will remain unsolved
Unfortunatelly, there are such cases, bur we always investigate them and adopt measures, such as trying to find a solution in cooperation with our clients, assinging new administrators to support the project, or offering a compensatory relief.
Most common causes when an issue is not being solved:
– in the very beginning, the business task was not described correctly, and we do not know about the problem;
– in the client's office, there is no consent as of the essence of an issue, when a chief, or a project manager, or developers want to solve it differently. Our administrators tend to beleive those who they contact more often, meaning developers;
– the business task just can not be solved by our administrator. Even if they do their best configuring your server, inherited web-site code errors will still be there.
If there is such a problem, we still can find a solution in the course of a meaningful dialogue.
Work service order
Before we start delivering service, we will discuss tasks, current issues, and propose the best option possible.
Interested in:
By pressing the "Send" button you consent to personal data processing as described in our Privacy Policy \ Нажимая кнопку, вы соглашается на обработку ваших данных согласно политике конфиденциальности