Typical issues we’ve solved

From the point of view of our experience we can say that most of our clients had similar issues. Also, usually they had some bad service before.

Good SysAdmin, SRE or DevOps service is quiet: first steps pass, application runs and nothing special happens. Many clients stop on that point because they do not understand processes required to sit under the hood for things to keep running. They hope that the server and application will continue running without service.

Some of pitfalls they suddenly could discover:

• Access to the hosting provider or server is lost. Nobody knows where the passwords are.
• Client selects another Website vendor, but the server appears to be the old vendor’s property.
• Website is reported as spreading viruses. Can we restore the code? Do we have backups?
• Website is down because server’s free space ran out.
• Data has been lost due to hard disk failure, RAID wasn’t set up.
• What are those hosting provider invoice items for? Where is the server infrastructure documentation?
• Virtual machine reset doesn’t help to restore Website availability anymore.
• Website loading time became very poor because of shared hosting usage. How to transfer the site to a dedicated server?
• Website SSL certificate expired and nobody cared about it beforehand. And how to issue new one?
• Comrade sysadmin has not been responding to the messages for two weeks even though the site has urgent issues.

We’ve cultured our operation’s processes exactly to solve issues like that. We accumulate a lot of data about client infrastructure, mainly as code (IaC).

We provide that data to the client as a Git repository on request for the client to be aware of it’s infrastructure also.

Typical infrastructure we operate

Here are example client infrastructures we could operate and recommend using. From the smallest to the more complex ones.

All estimations of Ops Hours below are approximate and couldn’t be treated as fixed cost project Offer. They depend on actual client requests, software peculiarities etc. Other vendor prices and proposals (e.g. Hetzner dedicated servers) are referred for time of writing this article and change frequently.

VPS for one Website

Use case: client needs something to keep its Website working with the access for coders to update the code. The simplest solution is a VPS anywhere near his customers.

Approximate costs for 6 (six) months are:

Dedicated server with two Environments

Use case: client needs one good server that can fit its production site and also needs separate environment for the development. Usually we recommend Hetzner EX61-NVME for that case.

Approximate costs for 6 (six) months are:

Simple Dedicated Infrastructure

Use case: client needs independent infrastructure with own backup server, GitLab, SaltStack instances and three independent environments for the full development cycle (development, staging, production).

Approximate costs for 6 (six) months are: