This space is reserved to evolve Serverless Knowledge collaboratively.
Any Serverless Enthusiast is encouraged to contribute to
creating content.
We ask for your help in developing the knowledge and also engagement rules for the community.
This space recognizes that it does not comprehend the entire Serverless Community.
There are active Serverless Community members that
do not know this space or don't want to contribute to it.
The ones that take part in this space, respect that decision and hope that
they decide to join in the future.
We do not elect ourselves as Serverless Community representatives or spokespersons.
Our main goal is to provide an additional spot
for the community to engage and develop itself.
To contribute, and join the effort, please go to our GitHub repository
On September 30, Gustavo Tavares and
Renato Losio published a survey to get the community's understanding of Serverless
Databases.
The survey close its collection period on October 20 with 32 answers.
The date when the survey closed allowed us to get a snapshot of the community's understanding on that day. Unfortunately, it became obsolete just after with all the announcements that happened just after. Like:
But our goal wasn't to have always the most updated list of Serverless Databases in the market. Our goal was to create a place for the community to evolve their knowledge collaboratively. And for that, we want to use the survey results as a first step.
This page documents the results of the survey and provides a preliminary analysis of its answers. In the end, we propose a method to keep the knowledge evolving. We would love to have you with us on his journey.
In the survey, we asked the community what would be the principles that a Database should follow to be considered Serverless.
We
offered a few predefined choices and allowed adding new choices.
The offered choices were:
Based on the responses we have a clear cluster of principles that most of the respondents agrees on:
Some respondents mentioned additional principles, like:
Our analysis is that, according to the community, at this date, a Serverless Database must present the following characteristics, in order of importance:
(*) We consider accepted the principles that achieve at least 50% of agreement from the participants of the survey.
After trying to understand what would be a Serverless Database,
we asked the community to evaluate a few existent offers in the
market.
The respondents could affirm whether a specific offer is considered or not as a Serverless Database.
Instead of
classifying the offer, the respondents also could inform they don't know the offer that we suggested.
The first information that we can take from the answers is the level of knowledge about the offers.
And the offers provided by AWS
have the biggest level of knowledge from the part of the respondents.
A question that can be made, based on the answers is the possibility of a biased selection of the respondents.
We accept that the
places where the survey was announced tend to have persons more active on AWS than the other cloud providers.
We plan to address this
in the next steps of this project.
Regarding which databases are considered Serverless, we got the below results:
In a more detailed analysis, all mentioned databases, considering only the respondents that know each offer, are below the 50%
threshold.
That means that more than half of respondents consider the offers as a Serverless Database.
Finally, we asked about other Database Offers that could be considered Serverless. A few choices were mentioned:
In the same question, we also received a few suggestions that the local or remote file systems should be considered as Serverless Database
offers.
We want to discuss this additional suggestion at a later moment.
From a more generic perspective, we asked if the respondents were able to name a so-called Serverless Database on each of the main Cloud Public Providers (AWS, GCP, and Azure).
Regarding the existent offers, the two questions mentioned above tried to look at the current state.
But we also were curious about
how the community sees the future of Serverless Databases.
For that, we added a question to understand,
based on the experience and knowledge of the respondents,
Where they expect that
Serverless Databases will gain market.
The question accepted multiple answers.
It was answered by 90% of the respondents.
We also asked a more generic question about the future of Serverless Databases.
Finally, we moved our focus to the current experience of the respondents.
We asked about their goal for using Serverless Databases.
We offered two predefined choices and left space available for them to provide new ones that were not listed.
We got 29 responses to that question.
22 answered with our suggested goals and 7 with additional goals.
We evaluated the
additional goals and clustered them around two concepts.
In the survey, we had a Free Text question where the respondents could mention the biggest challenge of Serverless Databases based on their experience.
Per the nature of the free text answers, is hard to mention everything here.
But we went through all of them and clustered the
challenges.
It is not possible to have a direct one-to-one relationship between the answers and the challenges,
But we believe the results shared
here do provide a good summary of the respondents' answers.
According to the researchers' understanding, the main challenges are (in order of importance):
According to the answers, it is clear that the respondents recognize that new ways of using databases exist.
And when Serverless
Databases are mentioned, they are exclusively focused on these "new" ways (NoSQL, Managed Services, etc.).
But the answers indicate that the main part of the applications being developed today is still following the relational model.
And
for these applications and projects, moving to Serverless Databases opens questions that are not easy to be addressed.
Here we understood that the challenges are not specific to the people that are answering the survey.
But they are broader and involve
other developers, project stakeholders, security departments, and others.
That is something that needs to be addressed in a follow-up
survey.
We also used a free text question to try to grasp what opportunities the community sees with Serverless Databases.
Again, we
summarized and clustered the answers here:
Here we close the report.
The report gives us an overview of the community's understanding of Serverless Databases on October 20th,
2022.
But we want to keep developing the knowledge and improve a shared understanding about this and other topics.
And we want
to invite you to join us.
The way that we want to do it is by using GitHub features like Forks, Pull Requests, and Issues.
If you want to take part in this
effort, please go to the Serverless Community Discussions repository and join
us.
Best,
Gustavo & Renato