《SlashData:2023年软件持续交付状态报告(英文版)(40页).pdf》由会员分享,可在线阅读,更多相关《SlashData:2023年软件持续交付状态报告(英文版)(40页).pdf(40页珍藏版)》请在三个皮匠报告上搜索。
1、2ABOUT THE CONTINUOUS DELIVERY FOUNDATION The Continuous Delivery Foundation(CDF)is an open source foundation that seeks to improve the worlds capacity to deliver software with security and speed.We help you figure out your best path to becoming a high-performing software delivery organization while
2、 leveraging open source.The open source projects we nurture are used by millions of developers worldwide.For more information on the CDF or our projects please visit:https:/cd.foundation/3CDF MEMBERS-PREMIERCDF MEMBERS-GENERALCDF MEMBERS-END USERCDF MEMBERS-ASSOCIATE4Can I share data from this repor
3、t?1.License GrantThis report is licensed under the Creative Commons Attribution-NoDerivativesLicence 4.0(International).Put simply,subject to the terms and conditions of this license,you are free to:Share You can reproduce the report or incorporate parts of the report into one or more documents or p
4、ublications,for commercial and non-commercial purposes.Under the following conditions:Attribution You must give appropriate credit to SlashDataTM,and to the Continuous Delivery Foundation as sponsors of this report,and indicate if changes were made.In that case,you may do so in any reasonable manner
5、,but not in any way that suggests that SlashDataTMendorses you or your use.NoDerivatives you cannot remix or transform the content of the report.You may not distribute modified content.2.Limitation of Liability SlashDataTM,believes the statements contained in this publication to be based upon inform
6、ation that we consider reliable,but we do not represent that it is accurate or complete and it should not be relied upon as such.Opinions expressed are current opinions as of the date appearing in this publication only and the information,including the opinions contained herein,are subject to change
7、 without notice.Use of this publication by any third party for whatever purpose should not and does not absolve such third party from using due diligence in verifying the publications contents.SlashDataTMdisclaims all implied warranties,including,without limitation,warranties of merchantability or f
8、itness for a particular purpose.SlashDataTM,its affiliates,and representatives shall have no liability for any direct,incidental,special,or consequential damages or lost profits,if any,suffered by any third party as a result of decisions made,or not made,or actions taken,or not taken,based on this p
9、ublication.The analyst of the developer economy|formerly known as VisionMobileSlashData Copyright 2023|Some rights reserved5Nikita SolodkovSenior Market Research AnalystNikita is a multidisciplinary researcher with a particular interest in using data-driven insights to solve real-world problems.He h
10、olds a PhD in Physics and has over five years of experience in data analytics and research design.nikita.solodkovslashdata.coLiam DoddSenior Market Research AnalystLiam is a researcher with an interest in data analytics and its ability to impact society.He holds a PhD in Physics,and previously worke
11、d generating intelligence and insights for the European automotive market.liam.doddslashdata.coABOUT THE AUTHORS6TABLE OF CONTENTS8KEY INSIGHTSINTRODUCTION716WHO IS INTO DEVOPS?HOW HAS SOFTWARE DELIVERY PERFORMANCE EVOLVED?1037WHAT DRIVES SOFTWARE DELIVERY PERFORMANCEMETHODOLOGY237KEY INSIGHTS The a
12、doption of DevOps practices continues to increase,with 84%of developers in DevOps-related activities as of Q1 2023.Testing applications for security measures has become the second most popular DevOps-related activity,with 37%of developers including it as an activity they are involved in.There is no
13、clear indication that the velocity for code changes has increased over the last two and a half years among the general developer population.Despite the ubiquity of DevOps practices,the increase in complexity of projects may be counteracting the benefits to development velocity.The correlation betwee
14、n speed and stability metrics remains strong,where 30%of high performers for code changes lead time are also high performers for service restoration.The average number of DevOps-related technologies used by developers has remained stable over the last year at 4.5 technologies.There is a strong corre
15、lation between the number of DevOps technologies used by developers and their likelihood of being a top performer across all three performance metrics tracked:lead time for code changes,deployment frequency,and time to restore service.Using CI/CD tools is correlated with better software delivery per
16、formance across all metrics.While application security testing can improve development velocity,the way it is implemented is crucial.Automated and continuous testing can be more effective than manual testing,but it is not a universal case.While using CI/CD tools can improve development performance,a
17、n increasing number of self-hosted CI/CD tools used by DevOps practitioners does not lead to greater performance,possibly due to interoperability issues.INTRODUCTION01Q1 2023|The State of Continuous Delivery-The Evolution of Software Delivery Performance91.IntroductionMoreover,we will see how develo
18、pers software delivery performance has changed over time.Finally,we will explore how developers usage of multiple continuous integration and continuous delivery(CI/CD)tools contribute to software delivery performance,as well as how approaches to application security testing change delivery performan
19、ce.The findings in this report are based on data from SlashDataspast six Developer Nation surveys,which reached more than 125,000 respondents worldwide over two and half years,from Q3 2020 to Q1 2023.Continuous Delivery(CD)is a software development practice that enables developers to release small b
20、ut frequent software updates reliably and safely.Closely linked to the broader DevOps cultural movement,CD consists of a set of practices that aim to automate and streamline the software delivery process.These practices,in turn,allow developer teams to innovate faster by collecting regular user feed
21、back and prioritise the product features and fixes that matter.To what extent,really,have developers embraced CD practices and the DevOps culture to increase the effectiveness of their software development and release process?In this report,commissioned by the Continuous Delivery Foundation(CDF)and
22、authored by SlashData,we explore the current state of the CD developer ecosystem and how this has evolved over the past two and half years.We will look at what proportion of developers are involved in the broader DevOps space and how many work with CD technologies in particular.WHO IS INTO DEVOPS?02
23、112.Who is into DevOps?Q1 2023|The State of Continuous Delivery-The Evolution of Software Delivery PerformanceIn the last year,we added an additional option for developers to include testing their applications for security measures among their DevOps activities.This immediately rose to the second mo
24、st popular activity,with 37%of developers including it as an activity they are involved in.Outside of this new practice,the relative ranking of activities has remained consistent,but with an increase in adoption across the board.Monitoring software and infrastructure saw the largest increase in rela
25、tive adoption,rising by 4 percentage points in the last year,to 40%.With this,monitoring software and infrastructure performance remains as the most popular DevOps-related activity,with testing applications for security measures close behind.Using CI to build and test code changes is now the third m
26、ost popular activity,with 37%of developers involved.Nearly half(49%)of developers use either CI or CD,up from 47%in Q1 2022,but only a little more than one in five(22%)developers are using CI and CD approaches to automate all building,testing,and deployment of code to production.This corresponds to
27、a small increase of 2 percentage points over the past 12 months,which continues the trend of a gradual increase in the adoption of CI/CD.For the sixth consecutive iteration of SlashDatas biannual Developer Nation survey,we asked developers where they are involved in any of the activities that common
28、ly fall under the DevOps spectrum,such as Continuous Integration(CI),CD,and infrastructure monitoring.As of Q1 2023,only 16%of developers are not involved in any DevOps-related activities.This represents a 7 percentage point decrease from Q1 2022,indicating a continual increase in the adoption of pr
29、actices that aim to increase an organisations ability to deliver software at high velocity.Despite the increase in developers involved in DevOps,only 7%describe their role as DevOps engineer/specialist,the same percentage as in Q1 2022.This highlights that even as more developers adopt DevOps practi
30、ces,they do so without necessarily self-identifying as specialists.02122.Who is into DevOps?Q1 2023|The State of Continuous Delivery-The Evolution of Software Delivery PerformanceIn terms of organisation sizes,our data reveal that medium-sized businesses have the highest involvement in DevOps practi
31、ces(94%).However,organisations of other sizes have seen significant growth in terms of the proportion of developers undertaking DevOps activities.Both small businesses and large enterprises now have more than 90%of their developers involved in DevOps-related activities,up by 6 and 5 percentage point
32、s,respectively,from Q1 2022.Further,79%of freelancers are now using DevOps practices,up from 71%last year.As noted in previous reports,DevOps practices have been broadly adopted across all sectors of the software economy.In fact,all regions have seen an increased adoption within their sectors.The se
33、ctors that had the lowest adoption of DevOps-related activities in Q1 2022,games and desktop apps,have seen large increases in their usage.81%of those involved in game development are using DevOps practices currently,compared to 73%in Q1 2022.Meanwhile,the desktop apps sector(85%)has overtaken mobil
34、e apps(84%),a 7 percentage point increase since Q1 2022.02132.Who is into DevOps?Q1 2023|The State of Continuous Delivery-The Evolution of Software Delivery PerformanceMore than 80%of developers are now involved in DevOps activities%of developers(n=23,387)40%37%37%36%34%29%27%23%1%16%I monitor softw
35、are and/or infrastructure performanceI test my applications for security vulnerabilitiesI use continuous integration to automatically build and test mycode changesI approve code deployments to productionI use continuous delivery/deployment to automate my codedeploymentsI programmatically provision a
36、nd manage IT infrastructuresI create automated regression tests and/or validation checksI build CI/CD pipelinesOther DevOps related activitiesNone of the aboveWhich of the following development activities are you involved in?02142.Who is into DevOps?Q1 2023|The State of Continuous Delivery-The Evolu
37、tion of Software Delivery PerformanceMore than 90%of developers in industrial IoT are involved in DevOps%of developers involved in DevOps activities by software sector(n=23,387)92%91%91%90%89%89%88%88%87%85%85%84%81%8%9%9%10%11%11%12%12%13%15%15%16%19%Industrial IoTEmbeddedsoftwareApps/extensionsfor
38、 3rd partyecosystemsConsumerelectronicsdevicesBackend servicesAugmentedreality includingnon-developersVirtual realityincluding non-developersData scienceWeb apps/Software as aServiceMachinelearning/AIDesktop appsMobile appsGamesInvolvement in DevOps by software sectorInvolved in DevOpsNot involved i
39、n DevOps02152.Who is into DevOps?Q1 2023|The State of Continuous Delivery-The Evolution of Software Delivery PerformanceDevelopers at medium-sized businesses have embraced DevOps practices%of developers involved in DevOps activities by company size(n=16,987)79%92%94%93%21%8%6%7%FreelancerSmall busin
40、ess(2-50 employees)Medium-sized business(51-1,000 employees)Large enterprise(More than 1,000 employees)DevOps involvement by organisational sizeInvolved in DevOpsNot involved in DevOpsHOW HAS SOFTWARE DELIVERY PERFORMANCE EVOLVED?03173.How has software delivery performance evolved?Q1 2023|The State
41、of Continuous Delivery-The Evolution of Software Delivery PerformanceAmong developers at large enterprises(more than 1,000 employees)we began to see an increase in top performers for lead time for code changes to 21%in Q1 2022,which has since decreased to 16%.However,despite the percentage of top pe
42、rformers decreasing,we still see that 40%of developers have lead time changes of less than one week,the second highest since we began asking developers.For deployment frequency,we have seen a continual decrease in the proportion of developers who are top performers those with multiple deploys per da
43、y.However,this decrease has been small over the scale that we have been monitoring,which,while concerning,does not indicate a huge crisis.We have also seen an increase in those deploying between once per hour and once per week.This suggests that increased DevOps practice adoption may help developers
44、 improve their performance,but reaching the highest performance levels is still a greater challenge.For individuals and organisations to measure the effectiveness of CD efforts,a set of robust performance metrics are required.In our survey,we ask developers about their performance for three of the f
45、our DORA metrics1:lead time for changes,deployment frequency,and time to restore service.These are used to measure software delivery performance and are predictive of organisational performance.Within the general developer population,our data show no clear signs that the velocity for code changes ha
46、s increased over the last two and a half years.Over this period,the percentage of top performers those with lead times of less than one day has fluctuated between 13%and 17%and currently stands at 15%in Q1 2023.It is possible that the increase in DevOps practices has not yet trickled down to positiv
47、ely impact performance.Moreover,the ubiquity of DevOps practices may be increasing the complexity of projects developers are involved in,counteracting the benefits to development velocity.1Forsgren,N.,Humble,J.,Kim,G.2018.Accelerate:The Science of Lean Software and DevOps:Building and Scaling High P
48、erforming Technology Organizations.IT Revolution Press03183.How has software delivery performance evolved?Q1 2023|The State of Continuous Delivery-The Evolution of Software Delivery PerformanceThe speed and stability metrics remain strongly correlated,rather than one compromising the other.More than
49、 half of developers(56%)who are low performers on lead time for code changes are also low performers for service restoration time.At the other end of the spectrum,30%of those who are high performers for lead time are also high performers for service restoration.A further 45%of lead-time high perform
50、ers take between one hour and one day to restore service,indicating good stability,even if these developers are not top performers.There has been a steady deceleration in the time it takes developers to restore service across the five survey waves from Q3 2020 to Q3 2022.However,in Q1 2023 this tren
51、d has abated slightly.With more than a third of all DevOps practitioners(34%)taking longer than a week to restore service,the proportion of low performers has continued to trend upward over the last two and half years.Conversely,the proportion of top performers has continued to decrease.With 12%of D
52、evOps practitioners taking less than an hour to restore service,it is no longer dropping precipitously.03193.How has software delivery performance evolved?Q1 2023|The State of Continuous Delivery-The Evolution of Software Delivery PerformanceThe proportion of top performers for lead time for code ch
53、anges has returned to levels comparable to a year ago%of DevOps practitioners(Q3 2020 n=10,252|Q1 2021 n=7,814|Q3 2021 n=8,784|Q1 2022 n=9,640|Q3 2022 n=13,108|Q1 2023 n=13,048)34%35%37%34%39%37%29%29%29%29%29%27%20%20%20%21%19%22%17%16%14%16%13%15%Q3 2020Q1 2021Q3 2021Q1 2022Q3 2022Q1 2023Software
54、delivery performance-Lead time for code changesLess than one dayOne day to one weekOne week to one monthMore than one month03203.How has software delivery performance evolved?Q1 2023|The State of Continuous Delivery-The Evolution of Software Delivery Performance35%37%37%33%37%35%32%32%33%33%31%32%21
55、%20%19%23%22%23%12%11%11%11%9%10%Q3 2020Q1 2021Q3 2021Q1 2022Q3 2022Q1 2023Software delivery performance-Deployment frequencyMultiple deploys per dayOnce per hour to once per weekOnce per week to once per monthLess frequently than once per monthQ3 2021 marked a high point for deployment frequency pe
56、rformance%of DevOps practitioners(Q3 2020 n=10,119|Q1 2021 n=7,613|Q3 2021 n=8,619|Q1 2022 n=9,473|Q3 2022 n=12,912|Q1 2023 n=12,747)03213.How has software delivery performance evolved?Q1 2023|The State of Continuous Delivery-The Evolution of Software Delivery Performance29%31%32%30%37%34%17%17%20%2
57、1%21%21%37%36%34%35%30%32%17%16%13%13%11%12%Q3 2020Q1 2021Q3 2021Q1 2022Q3 2022Q1 2023Software delivery performance-Time to restore serviceLess than one hourOne hour to one dayOne day to one weekMore than one weekTime to restore service performance has been decreasing over the last two and a half ye
58、ars%of DevOps practitioners(Q3 2020 n=9,349|Q1 2021 n=7,221|Q3 2021 n=8,126|Q1 2022 n=8,927|Q3 2022 n=12,385|Q1 2023 n=12,250)03223.How has software delivery performance evolved?Q1 2023|The State of Continuous Delivery-The Evolution of Software Delivery PerformanceLead time for code changes and time
59、 to restore service are closely linked%of DevOps practitioners(n=11,616)Lead time for code changesLess than one dayOne day to one weekOne week to one monthMore than one monthTime to restore serviceLess than one hour30%14%11%5%One hour to one day45%40%34%21%One day to one week15%30%23%18%More than on
60、e week10%16%32%56%5pp below the average of all other cells within each row2.5 5pp below the average of all other cells within each row2.5pp around the average of all other cells within each row2.5 5pp above the average of all other cells within each row5pp above the average of all other cells within
61、 each rowWHAT DRIVES SOFTWARE DELIVERY PERFORMANCE04244.What drives software delivery performanceQ1 2023|The State of Continuous Delivery-The Evolution of Software Delivery PerformanceNext,we analyse if the number of these DevOps technologies used by developers correlates with improved software deli
62、very performance.Our motivation is to examine whether having developers with a greater breadth of exposure to different aspects of DevOps and technologies to support them helps to collectively drive performance.On the contrary,it may be beneficial for developers to have a narrow and specific focus o
63、r responsibility and therefore have fewer tools or technologies to manage.In our survey,we capture information on a broad range of DevOps-related technologies that developers use,ranging from tools for managing source code to tools for monitoring application performance.The average number of technol
64、ogies that DevOps practitioners use,of those listed,has remained stable over the last year,at 4.5 technologies,on average.No technology has seen significant growth in the proportion of developers using them.The largest increase was observed for those using application security testing technologies,g
65、rowing by 3 percentage points to 28%over the last year.However,some technologies have seen a drop in their usage rates.Notably,the use of self-hosted CI/CD tools has dropped from 32%in Q1 2022 to 23%in Q1 2023.4.1 DevOps technology usage04254.What drives software delivery performanceQ1 2023|The Stat
66、e of Continuous Delivery-The Evolution of Software Delivery PerformanceAccording to our data,47%of practitioners using a single technology belong to the low-performing groups for lead time for code changes,44%to low performance of deployment frequency,and 55%to the time to restore service low-perfor
67、mance group.On the contrary,those who are using 10 or more technologies are more likely to be in the top-performing group than the low-performing group for lead time and restoration time,with deployment frequency being roughly equal.The results of this analysis resoundingly support the breadth of de
68、velopers involvement with DevOps technologies being highly beneficial to development velocity.There is a strong correlation between the number of technologies used and their likelihood to be a top performer.Even more important is that this finding is true across all three metrics we track,indicating
69、 that the benefit of utilising a breadth of technologies benefits DevOps practitioners across all areas of development velocity.04264.What drives software delivery performanceQ1 2023|The State of Continuous Delivery-The Evolution of Software Delivery PerformanceThe usage rate of self-hosted CI/CD to
70、ols has seen a significant decrease in the last 12 months%of DevOps practitioners(Q1 2022 n=14,176|Q1 2023 n=19,774)5%12%19%20%23%32%26%26%26%27%25%34%37%35%34%41%38%5%14%15%20%20%23%23%25%26%26%27%28%31%31%32%34%38%39%None of the aboveGitOpsFeature flaggingArtifact repositoriesIncident managementCl
71、oud-based IDEsSelf-hosted CI/CD tools(e.g.Jenkins,TeamCity)Managed CI/CD services(e.g.Azure Pipelines,AWS CodePipeline)Configuration managementInfrastructure monitoringInfrastructure as CodeApplication security testingAgile project management toolsApplication performance monitoring/observabilityColl
72、aboration/knowledge-sharing toolsTest automation/managementSource control managementIssue trackingDevOps technologies used in the past 12 monthsQ1 2023Q1 202204274.What drives software delivery performanceQ1 2023|The State of Continuous Delivery-The Evolution of Software Delivery PerformancePractiti
73、oners using more DevOps technologies are increasingly likely to be top performers in lead time for code changes%of DevOps practitioners by number of DevOps technologies used(n=13,048)0%5%10%15%20%25%30%35%40%45%50%+Code lead time for changes against number of DevOps technologies usedNumbe
74、r of DevOps technologies usedLess than one dayMore than one month04284.What drives software delivery performanceQ1 2023|The State of Continuous Delivery-The Evolution of Software Delivery PerformanceThose using 10 or more tools are more than twice as likely to be high performers than those only usin
75、g a single technology%of DevOps practitioners by number of DevOps technologies used(n=12,746)Number of DevOps technologies usedMultiple deploys per dayLess frequently than once per month0%5%10%15%20%25%30%35%40%45%+Deployment frequency against number of DevOps technologies used04294.What
76、drives software delivery performanceQ1 2023|The State of Continuous Delivery-The Evolution of Software Delivery PerformanceA greater breadth of DevOps tools used is associated with dramatically fewer low performers for service restoration%of DevOps practitioners by number of DevOps technologies used
77、(n=12,249)Number of DevOps technologies usedLess than one hourMore than one week0%10%20%30%40%50%+Time to restore service against number of DevOps technologies used04304.What drives software delivery performanceQ1 2023|The State of Continuous Delivery-The Evolution of Software Delivery Pe
78、rformanceDevelopers who use CI/CD tools are more likely to be top performers compared to those who do not.This is particularly stark for top performers in time to restore service,where those who use CI/CD tools(17%)are more than twice as likely to be top performers than those who do not(8%).However,
79、what may be of more importance is the likelihood that a practitioner will be in the low-performing groups.For those not using CI/CD tools,the low-performing groups are the largest proportion of developers for all metrics.The reverse is true for those using CI/CD tools,where the low-performing group
80、is not the largest group for any of the metrics and is substantially smaller than those who do not use CI/CD tools.While the breadth of technology usage has been shown to be beneficial to development velocity,certain tools are found to be more impactful than others.When comparing DevOps practitioner
81、s who use CI/CD tools managed or self-hosted against those who do not,we can see a distinct difference in performance across all three development metrics.4.2 CI/CD and application security testing04314.What drives software delivery performanceQ1 2023|The State of Continuous Delivery-The Evolution o
82、f Software Delivery PerformanceWhile application security testing can be associated with increased development velocity,it is not a simple binary switch,nor does automating the processes universally help developers.Those who are introducing application security testing measures should carefully cons
83、ider their uses and ensure they operate well within their current processes,otherwise they may not benefit from development velocity as intended.The way various DevOps tools are used also has an impact on performance.In our Q3 2022 survey,we asked DevOps practitioners who practise application securi
84、ty testing what kind of testing they did.Further,we asked them if it was automated and continuous,manual,or ad-hoc.In the three examples shown,we can see different effects of this in practice.Developers who perform build-time security checks in an automated and continuous fashion are the most likely
85、 to be top performers,and the least likely to be low performers,across all three metrics,of the types shown.Contrast this against those who perform this testing manually,who are the worst performers across the three metrics.Similarly,for developers performing test-time security checks,those using au
86、tomated testing are slightly more likely to be top performers than those doing so manually.However,they are several percentage points less likely to be in the low-performing groups.04324.What drives software delivery performanceQ1 2023|The State of Continuous Delivery-The Evolution of Software Deliv
87、ery PerformanceThose who use CI/CD tools are significantly less likely to be low performers than those who do not%of DevOps practitioners for each performance metric(Lead time for code change n=13,048|Deployment frequency n=12,747|Time to restore service n=12,250)13%20%25%43%10%22%30%39%8%27%23%42%1
88、7%25%30%28%11%25%35%29%17%41%18%24%Less than onedayOne day to oneweekOne week toone monthMore than onemonthMultipledeploys perdayOnce per hourto once perweekOnce per weekto once permonthLess frequentlythan once permonthLess than onehourOne hour toone dayOne day to oneweekMore than oneweekPerformance
89、 against CI/CD usageNot using CI/CDUsing CI/CDLead time for code changeDeployment frequencyTime to restore service04334.What drives software delivery performanceQ1 2023|The State of Continuous Delivery-The Evolution of Software Delivery Performance10%13%11%12%12%11%45%36%43%40%39%40%Performed manual
90、ly/ad-hocAutomated and continuousPerformed manually/ad-hocAutomated and continuousPerformed manually/ad-hocAutomated and continuousLead time for code changes Automated build-time security checks see the greatest proportion of top performers among the highlighted application security measures%of DevO
91、ps practitioners for each performance metric(Lead time for code change n=12,605|Deployment frequency n=12,469|Time to restore service n=12,062)9%13%10%12%11%12%42%34%42%37%37%38%Time to restore serviceLow performersTop perfomers7%9%8%8%9%9%42%33%39%35%37%36%Deployment frequencyAPI security testingTe
92、st-time security checksBuild-time security checks04344.What drives software delivery performanceQ1 2023|The State of Continuous Delivery-The Evolution of Software Delivery PerformanceWe note that the proportion of top performers remains flat while that of low performers increases dramatically,with a
93、n increasing number of self-hosted CI/CD tools used.This suggests that there is a diminishing return from increasing the number of CI/CD tools a developer uses.The usage of an increasing number of tools may also be a response to increased complexity,which is having negative impacts on the performanc
94、e of these developers.Similarly,the integration of multiple tools may not be optimally implemented,leading to function overlap that is impacting performance.We have previously shown that using CI/CD tools increases the likelihood of DevOps practitioners being top performers across lead time for code
95、 changes,deployment frequency,and time to restore service.However,there are noted concerns that interoperability issues may emerge when multiple CI/CD tools are used in conjunction with one another.Across all three of the development performance metrics used,we can see that an increased number of se
96、lf-hosted CI/CD tools used is not associated with greater performance.For lead time for code changes and deployment,we see minimal increases in the likelihood of DevOps practitioners being high performers as they increase the number of CI/CD tools used.For both,those using five CI/CD tools see the l
97、owest percentage of top performers,uplifting for those using six or more,but not to levels above those using fewer CI/CD tools.4.3 Interoperability of CI/CD tools04354.What drives software delivery performanceQ1 2023|The State of Continuous Delivery-The Evolution of Software Delivery PerformanceAn i
98、ncreasing number of tools used having such a strongly negative impact on service restoration time has multiple possible explanations.However,interoperability issues may be at the centre of many of them.Multiple tools may make it challenging to integrate all of them well,leading to a greater challeng
99、e to isolate the service-impacting issue at hand.Further,a lack of standardisation between tools may make it more difficult for all tools to work together well,which also increases the challenge of addressing service failure.The time to restore service metric sees the most dramatic increase in low p
100、erformers from the increased number of self-hosted CI/CD tools used.Among those using one self-hosted CI/CD tool,the likelihood of being a top performer is greater than being a low performer.However,once a practitioner passes two or three CI/CD tools,this reverses dramatically.Among developers who u
101、se more than five tools,more than half are low performers,and less than 10%are top performers.04364.What drives software delivery performanceQ1 2023|The State of Continuous Delivery-The Evolution of Software Delivery PerformanceUsing multiple self-hosted CI/CD tools can lead to long response times t
102、o service failure%of DevOps practitioners for each performance metric(Lead time for code change n=13,048|Deployment frequency n=12,746|Time to restore service n=12,249)0%10%20%30%40%50%60%123456+Lead time for code changes performance against the number of CI/CD self-hosted tools usedLess than one da
103、yMore than one monthNumber of CI/CD self-hosted tools used0%10%20%30%40%50%60%123456+Deployment frequency performance against the number of CI/CD self-hosted tools usedMultiple deploys per dayLess frequently than once per month0%10%20%30%40%50%60%123456+Time to restore service performance against th
104、e number of CI/CD self-hosted tools usedLess than one hourMore than one weekMETHODOLOGYThe Developer Nation SurveyThe 24th edition of the Developer Nation survey reached nearly 26,000 respondents from 166 countries around the world.As such,the Developer Nation series of surveys continues to be the m
105、ost global independent research on mobile,desktop,industrial IoT,consumer electronics,embedded,third-party app ecosystems,cloud,web,game,augmented and virtual reality,and machine learning developers and data scientists combined,ever conducted.The report is based on a large-scale,online developer sur
106、vey designed,produced,and carried out by SlashData over a period of ten weeks between December 2022 and February 2023.Respondents to the online survey came from 166 countries,including major app and machine learning development hotspots such as the US,China,India,Israel,the UK,and Russia,even stretc
107、hing all the way to Kenya,Brazil,and Jordan.The geographic reach of this survey is truly reflective of the global scale of the developer economy.The online survey was translated into eight languages in addition to English,namely simplified Chinese,traditional Chinese,Spanish,Portuguese,Vietnamese,Ru
108、ssian,Japanese,and Korean,and was promoted by more than 100 leading community and media partners within the software development industry.To eliminate the effect of regional sampling biases,we first weighted to correct for over-represented individual countries within regions.We then weighted the reg
109、ional distribution across nine regions by a factor that was determined by the regional distribution and growth trends identified in our Developer Nation research.Each of the separate branches:mobile,desktop,Industrial IoT,consumer electronics,embedded software,third-party app ecosystems,cloud,web,ga
110、mes,augmented and virtual reality,and data science and machine learning were weighted independently and then combined.To minimise other important sampling biases across our outreach channels,we weighted the responses to derive a representative distribution for technologies used and developer segment
111、s.Using ensemble modelling methods,we derived a weighted distribution based on data from independent,representative channels,excluding the channels of our research partners,to eliminate sampling bias due to respondents who were recruited via these channels.Again,this was performed separately for eac
112、h of mobile,industrial IoT,consumer electronics,embedded software,third-party app ecosystems,desktop,cloud,web,games,augmented and virtual reality,and data science and machine learning.For more information on our methodology please visitSlashData Website:Our methodology pageA short video explaining
113、our methodology WHO DEVELOPERS AREWHERE THEY ARE GOINGWHAT THEY BUYDeveloper population sizingDeveloper segmentationWe help the world understand developersWhy developers are adopting competitor products and how you can fix thatEmerging platforms augmented&virtual reality,machine learningWe survey 30
114、,000+developers annually across Web,Desktop,Cloud,Mobile,Industrial IoT,AR/VR,Machine Learning and Data Science,Games,Consumer Electronics and Apps/Extensions for 3rd party ecosystems-to help companies understand who developers are,what they buy and where they are going next.39TRUSTED BYthe leading tech platformsO U R C L I E N T STHE ANALYST OF THE DEVELOPER ECONOMYwww.slashdata.co