Each year Ceph holds a user survey to measure trends, improvements, and actual requests for functionalities. These surveys are very convenient for making hypothesizes about user behaviors as well. Thus, each year new user surveys are conducted to keep the data up to date. Two weeks ago, I shared some highlights about the 2021 OpenStack user survey which can be found here: https://www.fairbanks.nl/some-highlights-of-the-openstacks-2021-user-survey/ . In this blog some highlights of the 2021 Ceph user survey.
Mainly, the results reveal that in using Ceph, the value of open-source and scalability have consistently remained among the top priorities, and when users are allowed to select multiple usage considerations then high availability becomes the third most important characteristic.
The next noticeable thing is that Ceph users report deployments across all continents except Antarctica (for now!).
Also, when the survey participants were asked “Where do you think the Ceph developer community should focus its efforts?” and given a pre-defined set of responses, users rank priorities as seen in the graph below, with ‘reliability’ as the clear top priority and ‘performance’ and ‘documentation’ tied for second.
Users were also asked to identify their top wish list features which could be seen as another indicator of priorities. In this case, rather than asking users to prioritize feature categories, they were asked to provide a short qualitative response. The results of this are viewable in the graph below.
These results indicate that users are comfortable with the current level of reliability because no feature requests fell into that category, consistent with the NPS responses but inconsistent with the weighted category prioritization. Furthermore, ‘Ease of use’ features appeared as the second-most commonly requested feature.
Users seem generally satisfied with the current levels of performance. It could be argued that this is related to an organizational trend to consolidate more workloads or storage service tiers into fewer disparate storage systems.
Moreover, the majority of users reported running Nautilus or earlier versions (71%), which does not match users’ stated upgrade behaviors, with 63% claiming to update major releases within a year or sooner after the release is available. While one might assume that users would run Octopus side-by-side with an earlier release to test prior to rolling the new version, only 19% of users report doing so.
Moving on, 23% of users reported using their clusters exclusively for one storage interface, demonstrating that users appreciate the versatility of multiple storage interfaces from a single cluster. This number is reinforced by the fact that 34% of users reported a single-use purpose for their Ceph cluster, the other 66% stated using the cluster for at least two distinct use cases.
Those were the biggest highlights of the Ceph User Survey of 2021! Did you notice some highlights as well that are not mentioned in this blog, or are your Ceph usage and requirements completely different? Let us know in the comments and stay tuned for our blog about the Ceph User Survey results of 2022!
More details about the survey can be found here: https://ceph.io/en/news/blog/2021/2021-ceph-user-survey-results/