Analysis of 2022 User Survey Feedback

Introduction

This is an analysis of the results from the 2022 User Survey in comparison to the 2021 results. As with the previous analysis, the analysis will be performed on a per-question basis with an over-all summary at the end.

How do you upgrade your version of OpenStack?

Response

Users

Percentage of Responses

Skip/fast-forward releases

84

34

Upgrade all coordinated releases (once every 6 months)

75

30

Not upgrade

65

26

Deploy regularly from the master branch

12

4

Deploy all intermediary releases and all coordinated releases

12

4

The total number of responses was slightly lower than in the 2021 survey - 272 in 2021 and 249 in 2022. Results in this year’s survey are almost the same as from last year. The only differences are a small increase of users who do not do upgrades and in the same time decrease amount of users who deploy all intermediary releases and all coordinated releases. But as differences are small (2-3%) this doesn’t show any significant change. It is very interesting how results of this question will change (or not) in next year’s survey, when SLURP releases will be available.

Once on a given release, do you use stable branches for bug-fix upgrades?

Number of responses for this question was a bit smaller than in 2021, similar to the previous questions. There were 235 responses in 2022 compared to the 265 in the 2021 survey.

Response

Users

Percentage of Responses

I do not do bugfix upgrades

39

17

Yes, backporting specific fixes

50

21

Yes, deploying every commit on the stable branch

16

7

Yes, upgrading at various points in time depending on fixes

71

30

Yes, using only official point releases

59

25

The responses for the question about stable branches updates are very consistent with 2021 survey. There are no major changes in how results are distributed. The only change is an increase in the number of users who are backporting specific fixes (21% this year vs. 17% in 2021) and a decrease in the number of users who are using only official point releases (25% this year vs. 30% in 2021). This shows that there is a significant number of users who are backporting fixes selectively on their own rather than waiting for official point releases which will include required fixes.

To which projects does your organization contribute maintenance resources such as patches for bug and reviews on master or stable branches?

Overall participation level for this questions is similar to the 2021 survey - about 44% of users who provided information about used projects, participated also in this question. In the TOP 5 projects according to the number of contributors there is one change - Ironic is now listed as 5th project with top number of contributors. It replaced Kolla which was ranked 5th in the 2021 survey. Majority of the projects (3/5) showed minor decrease in reported number of contributors from 2021.

Project

Contributors

Contributors in 2021

Nova

63

63

Neutron

53

56

Cinder

31

40

Keystone

30

34

Ironic

25

23

After the first review of these results in the 2019 user survey, the TC wanted to compare the number of respondents who reported they used a project as compared to the number who reported that they contribute to the project. For the 2021 results review we have continued the analysis to see how many of respondents who reported they used a project also contribute to the project. In 2022 we have continued that but have also added the previous percentage of participation for reference. To get this information we considered the number of users who reported that they were using the project in production. Note that we did not count users who indicated that they were just testing a service or who indicated that they had the service installed as part of a Proof of Concept. So the number of users that are using any given service may be notably larger than indicated in the results below. Here are the results of that investigation.

Project

Contributors

Users

% Participation

2021 % Participation

AODH

3

28

11

9

Barbican

10

81

12

11

Blazar

3

5

60

150

Ceilometer

10

75

13

12

Cinder

31

213

15

19

Cloudkitty

5

12

42

33

Congress

2

0

Cyborg

2

7

29

75

Designate

14

53

26

24

Freezer

2

0

Glance

21

218

10

12

Heat

16

154

10

8

Horizon

18

199

9

8

Ironic

25

64

39

35

Keystone

30

226

13

15

Kolla

18

59

31

49

Kuryr

2

5

40

63

LOCI

3

6

50

43

Magnum

10

50

20

42

Masakari

5

8

63

14

Manila

10

36

28

50

Mistral

6

16

38

35

Monasca

5

0

12

Murano

1

8

13

50

Neutron

53

216

25

24

Nova

63

217

29

27

Octavia

21

112

19

29

OpenStack-Ansible

23

58

40

46

OpenStack-Helm

4

8

50

36

OpenStack Client

19

179

11

11

Panko

10

0

43

Qinling

0

0

Rally

5

38

13

14

Sahara

1

2

50

50

Searchlight

1

0

Senlin

4

4

100

80

Solum

0

0

Storlets

0

0

Swift

18

90

20

19

Tacker

1

1

100

33

Tricircle

1

1

100

TripleO

3

23

13

42

Trove

6

9

67

29

Vitrage

2

2

100

Watcher

2

3

67

33

Zaqar

5

0

22

Zun

4

0

25

There are few things worth to note regarding the list above. Projects like: Monasca, Panko, Zaqar and Zun still have users who use it in the Production but they lost all of their contributors. Those projects joined the group of the projects with 0 declared contributors which includes also: Congress, Freezer, Solum and Storlets. Couple of projects like TripleO, Murano or Cyborg have experienced significant drop in the participation level in 2022 comparing to the previous year. The good thing to note is that none of the core projects like e.g. Cinder, Keystone, Glance, Ironic, Neutron, Nova, Horizon, OpenStack Client have experienced contributors drop.

How do members of your organization contribute to OpenStack?

As with the other questions in the user survey, this question also saw an decrease in responses. 111 users responded to it in this years’ survey comparing to the 173 users last year.

Response

Users

Percentage of Responses

Bug reports

93

84

Documentation improvement

47

42

Participate in Forum sessions at the Summit

44

40

Bug fixes on master

43

39

Participate in Ops meetups

37

33

Participate in PTG sessions

30

27

Code review on master

27

24

Backporting bug fixes to stable branches

25

23

Sponsor in-person events

24

22

Code review on stable branches

24

22

Feature development

23

21

Feature design review

14

13

Contribute resources to run CI jobs upstream

8

7

Host third-party CI jobs downstream

7

6

Other

1

1

Similarly to the last year results, most people declare that they report bugs as their way of contribution to the OpenStack project. Very good thing to note is the fact that “Documentation improvement” response have got significant increase of the users (47 users this year vs. 39 in the 2021). From the other hand different ways of the participation in the events decreased significantly. One of the potential reasons for that for the in person events may be effect of the overall global economic situation where many companies have experienced budget difficulties and are looking for a ways how to save some money. Other potential reasons for that may be for example:

  • health concern,

  • environmental concerns,

  • political issues.

This is just a speculation as there was no question about this in the survey. If we want to know more details about what prevents people from participating in the in person events, We should consider adding new question to the next user survey.

What prevents you or your organization from contributing more maintenance resources, or makes contributing difficult?

There were 101 users who provided response for this question in the 2022 survey. It is a decrease of the number of responses comparing to the 2021 where 131 users answers to this question.

Response

Users

Percentage of Responses

Lack of resources

54

53

Lack of skills

14

14

Bug tracker tool is not intuitive

11

11

Inconvient way of receiving patches

11

11

Difficult process

8

8

Business decission

3

3

Using too old OpenStack version

3

3

No need to contribute as it just works as needed

3

3

Paying vendor for support

2

2

Specific changes, rejected by the community

2

2

Language barrier

2

2

Security; Private Network

1

1

Company focus shifted to other projects

1

1

Difficult to get started

1

1

The most common reason which prevents organizations from contributing is lack of resources. The second from the top, also repeated many times was lack of skills. There was also one user who responded that it’s “difficult to start” with them which could actually be included in the same category.

There was 8 people who replied that the main reason which stops it from contributing is generally “difficult process”. It’s worth to look deeper into this category of replies and see in detail what were the answers there:

Most problems we encounter are in the gaps between projects which means neither project team will take responsibility and review/help out. “Drive by” patches from deployers are often one liners and attract -1 reviews on the basis of “theres no testing of this whole area in devstack- so this is -1”. Operators have little or no interest in devstack and and not in a position to fix it’s technical debt.

Some projects have major lead times between patch contributions and reviews, with many picky reviews on non-critical areas. Thus we are not able to contribute significantly to Nova for example, unless needed because the staff get frustrated with the lack of progress and are on short contracts.

Cumbersome upstream review processes make contributing fixes a time and thus cost intensive undertaking - this is the reason none of our drivers was ever tried to put to upstream but rather made public ourselfes.

Internal struggle with OS policies.

scope creep in bugfixes/regressions (i.e. “please make your re-added feature which fixes a regression generally usable for ipv4 and ipv6” when the original solution- which was implemented upstream- didn’t implement ipv6 at all)

The time to have a fix “production ready” for upstream.

The onboarding and contribution process is pretty high-bar (custom infrastructure compared to e.g. gitlab.com/github.com, gerrit workflow is unfamiliar to many); Oftentimes the issues we encounter are already fixed upstream.

From above replies it seems that main reason why users see our process as difficult is generally review process and requirement for high quality patches, with good testing coverage when many operators just wants to have their issue fixed fast.

Other interesting point is 3 users who replied that they don’t need to contribute as OpenStack works for them as expected.

Other ways users participate:

There were no users responses to this question this year. It’s very similar to what was in the 2021 survey where we had just 2 responses to the same question.

Summary

Unfortunately we noticed decrease of the overall number of users responding to the TC questions in 2022 user survey comparing to the 2021 survey. The good thing is that there is still a significant number of responses so we can get a lot of useful information from it. After 2021 survey numbers in most cases are very similar in the 2022 survey. It seems that things like, how users are upgrading OpenStack, what maintenance releases they are using or how people are contributing to the OpenStack in general is stable.

The fact that there is still increasing number of projects with lack of contributors is alarming and TC should probably thing more about actions possible to take to address it.

Additional Resources

The OpenStack Survey Report also provides a graphical overview of the OpenStack Survey results.