1、1NOVEMBER 2020 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,

2、000+developers annually across web,mobile,IoT,cloud,Machine Learning,AR/VR,games and desktop to help companies understand who developers are,what they buy and where they are going next.TRUSTED BYthe leading tech platformsO U R C L I E N T S4Can I share data from this report?1.License GrantThis repor

3、t 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 publications,for commercial

4、and non-commercial purposes.Under the following conditions:Attribution You must give appropriate credit to SlashDataTM,and to Sentry,as sponsors of this report,and indicate if changes were made.In that case,you may do so in any reasonable manner,but not in any way that suggests that SlashDataTMendor

5、ses 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 information that we consider reliable,but we do not represent

6、 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 without notice.Use of this publication by any third pa

7、rty 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 fitness for a particular purpose.SlashDataTM,its affilia

8、tes,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 publication.The analyst of the developer economy|formerl

9、y known as VisionMobileSlashData Copyright 2022|Some rights reservedTable of contents5Q3 2022|INFRASTRUCTURE VS APPLICATIONS REPORT FOR SENTRYKey insights|61.Introduction|72.DevOps and SRE activities|93.Development culture and KPIs|164.The cost of application issues/infrastructure outages|22 Appendi

10、x|276Key insights62%of SRE/DevOps practitioners report that the most common challenge they face is an unclear ownership boundary between application and infrastructure issues.The number one activity that SRE/DevOps specialists would like to spend less time doing is chasing down others to resolve app

11、lication issues,with 74%of respondents agreeing to this statement.SRE/DevOps specialists are more likely to spend time fixing application code than vulnerabilities in infrastructure nearly three in ten spend 40%+of their time fixing application issues,compared to nearly two in ten who spend 40%+of t

12、heir time on infrastructure.30%of respondents estimate that their SRE/DevOps teams lose more than one person-month in productivity per year fixing application issues.Being proactive is the primary goal for most respondents;64%indicate that they would like to spend more time proactively analysing/fix

13、ing vulnerabilities in software infrastructure.More than 76%of SRE/DevOps practitioners would like to spend more time working with development teams to run tests for scaling up software,providing input to system design decisions,and increasing automation of development processes.52%of SRE/DevOps spe

14、cialists work in organisations where the engineering culture is“you build it,you own it”.58%of SRE/DevOps practitioners would prefer departments to have more autonomy and control over their tooling decisions.12%of respondents estimate that their organisation has had to pay out more than$100K(USD)in

15、compensation in the past year for service-level agreement violations as a result of application issues.1.Introduction81.IntroductionQ3 2022|INFRASTRUCTURE VS APPLICATIONS REPORT FOR SENTRYIn this report,we provide data and insights from a global survey of 141 Site Reliability Engineers(SREs)and DevO

16、ps practitioners commissioned by Sentry and fielded in Q3 2022 by SlashData.All respondents have SRE or DevOps responsibilities within organisations with more than 50 employees and live software applications in production.SREs and DevOps practitioners play a critical role in software development org

17、anisations,ensuring that software systems are ultra-scalable and highly reliable and that updates can be delivered quickly and effectively with low risk or cost of failure.However,almost two-thirds(62%)of SRE/DevOps practitioners face significant challenges in determining boundaries between ownershi

18、p of application code and infrastructure and in managing blurred SRE/DevOps and developer responsibilities.Around half(52%)of SRE/DevOps practitioners currently work in organisations with a culture of developers managing the performance of code that they build,both to create clearer ownership bounda

19、ries and to enable SRE and DevOps practitioners to focus more on proactive work to manage and scale infrastructure.We expect to see this kind of decentralisation becoming even more common in future.While SRE requires a significant focus on ensuring the continuous operation of infrastructure,many who

20、 specialise in this role find themselves balancing other responsibilities.We show that SRE and DevOps practitioners spend more time resolving application code issues compared to infrastructure issues,and that the top activity theyd like to spend less time doing is chasing down others to resolve them

21、.Being proactive is a key goal for most respondents,with the majority indicating that they would like to spend more time proactively analysing/fixing vulnerabilities in software infrastructure.In chapter 2 of this report,we look at which activities SRE/DevOps practitioners are spending most of their

22、 time on,as well as which activities they would like to spend more or less time on to improve the effectiveness of SRE and DevOps within their organisation.In chapter 3,we explore the cultures that exist in terms of ownership and management of performance,the autonomy of development teams to decide

23、their tooling,as well as the KPIs that SRE/DevOps teams are currently using to manage software performance.Finally,in chapter 4,we present the current impact of application issues and infrastructure outages in terms of costs and damages incurred and loss of productivity.2.DevOps and SRE activities10

24、2.DevOps and SRE activitiesQ3 2022|INFRASTRUCTURE VS APPLICATIONS REPORT FOR SENTRY83%of SRE/DevOps practitioners report that they spend some proportion of their time analysing and fixing vulnerabilities in application code.Thats considerably more than those who spend time proactively addressing vul

25、nerabilities in the behind-the-scenes infrastructure that supports a software system and helps to make development more efficient(63%).In fact,fixing vulnerabilities in application code frequently accounts for a significant proportion of an SRE/DevOps specialists time 29%of respondents report that 4

26、0%or more of their working time is spent in this way.112.DevOps and SRE activitiesQ3 2022|INFRASTRUCTURE VS APPLICATIONS REPORT FOR SENTRYSRE/DevOps practitioners are more likely to spend time fixing application code than vulnerabilities in infrastructure%of professional SRE/DevOps practitioners(Q3

27、2022 n=141)19%36%39%37%36%24%17%19%17%11%55%37%33%33%33%45%44%42%39%34%10%12%11%12%9%11%15%21%15%24%5%5%6%7%12%8%12%9%14%13%6%6%6%7%7%7%5%8%10%6%4%5%10%3%5%5%4%8%8%Responding to incidents/outages in infrastructureRunning tests to scale-up software/systemsEquipping development teams with toolsAnalysi

28、ng/fixing vulnerabilities in infrastructureBuilding tools for engineers/developers to interact with infrastructureToil(routine tasks to maintain service or production environment)Automating development processes/writing and maintaining automation codeChasing down others to resolve application issues

29、(e.g.errors/crashes,performance)Analysing/fixing vulnerabilities in application codeProviding input to software/system design decisionsHow much of your time is spent on each of the following activities?I dont do thisLess than 20%of my time20%to 39%of my time40 to 59%of my time60%to 79%of my time80%o

30、r more of my time122.DevOps and SRE activitiesQ3 2022|INFRASTRUCTURE VS APPLICATIONS REPORT FOR SENTRYWe also asked SRE/DevOps practitioners which activities they felt those within their organisation should spend more or less time on to manage the performance and delivery of software systems more ef

31、fectively.The number one activity that SRE/DevOps specialists would like to spend less time doing is chasing down others to resolve application issues,with 74%of them agreeing to this statement.Moreover,64%of respondents indicated that they would like to spend more time proactively analysing/fixing

32、vulnerabilities in software infrastructure.Between 76%and 82%of respondents would like to have more time to dedicate to supporting development teams by running tests to scale up software,providing input to system design decisions and increasing automation of development processes.132.DevOps and SRE

33、activitiesQ3 2022|INFRASTRUCTURE VS APPLICATIONS REPORT FOR SENTRYThe majority of respondents would like SRE/DevOps practitioners within their organisation to spend less time chasing others to resolve application issues and more time proactively addressing infrastructure vulnerabilities%of professio

34、nal SRE/DevOps practitioners(Q3 2022 n=141)26%33%40%64%70%72%76%79%82%74%67%60%36%30%28%24%21%18%Chasing down othersto resolve applicationissues(e.g.errors/crashes,performance)Toil(routine tasks tomaintain service orproductionenvironment)Responding toincidents/outages ininfrastructureAnalysing/fixin

35、gvulnerabilities ininfrastructureBuilding tools forengineers/developersto interact withinfrastructureEquippingdevelopment teamswith toolsRunning tests toscale-upsoftware/systemsProviding input tosoftware/systemdesign decisionsAutomatingdevelopmentprocesses/writingand maintainingautomation codeWhich

36、of the following activities do you feel your organisation should spend more/less time on in order to perform DevOps or SRE more effectively?More timeLess time142.DevOps and SRE activitiesQ3 2022|INFRASTRUCTURE VS APPLICATIONS REPORT FOR SENTRYWe also asked respondents to indicate their main challeng

37、es concerning their SRE/DevOps responsibilities.The most commonly reported challenge is unclear boundaries between the developer and SRE/DevOps roles and between ownership of the application and infrastructure-related issues.This shows that a culture of clearer ownership and responsibility is needed

38、,as well as a shift in how organisations structure their teams.For example,having developers who build application code also managing application performance,while SRE/DevOps practitioners focus on proactive work to manage aspects such as infrastructure reliability,tooling and delivery.152.DevOps an

39、d SRE activitiesQ3 2022|INFRASTRUCTURE VS APPLICATIONS REPORT FOR SENTRYUnclear boundaries between application and infrastructure ownership or SRE and developer roles are the most common challenges among SRE/DevOps practitioners%of professional SRE/DevOps practitioners(Q3 2022 n=141)62%37%27%24%22%1

40、9%13%11%Unclear boundariesbetween applicationand infrastructureownership,orSRE/DevOps anddeveloper rolesTeams are using toomany different toolsHaving inadequatetools to monitoroverall performanceof your applications,or identify and fixerrors quicklyUnclear resolutionpath for softwareapplication issu

41、esPressure toconsolidate toolingfor cost-savingNot having theautonomy to choosethe right tooling foryour jobSub-teams not havingthe autonomy topurchase the righttools for their jobsNone of aboveMain challenges in relation to SRE/DevOps activities3.Development culture and KPIs 173.Development culture

42、 and KPIs Q3 2022|INFRASTRUCTURE VS APPLICATIONS REPORT FOR SENTRYOur data shows that more than half of all respondents work for organisations where ownership and performance management of software are decentralised i.e.those who build software also own and manage the performance of their applicatio

43、ns.Another third of respondents work for organisations where SRE/DevOps practitioners manage the performance of applications created by developers.However,as we have seen in our earlier findings,this often leads to challenges around blurred ownership and responsibility,as well as taking time away fr

44、om SREs/DevOps key activities.For modern organisations to operate even more efficiently,we expect to see the culture of decentralisation becoming even more common in future.More than two-thirds of respondents currently work for organisations where departments have the autonomy to choose their toolin

45、g,irrespective of whether they have their own budget or not.Our results show that SRE/DevOps practitioners clearly prefer departments to have more autonomy and control over their tooling decisions rather than budgets and decision-making being centralised.183.Development culture and KPIs Q3 2022|INFR

46、ASTRUCTURE VS APPLICATIONS REPORT FOR SENTRYJust over half of respondents work in organisations where those who build software are also responsible for managing performance in production%of professional SRE/DevOps practitioners(Q3 2022 n=141)52%32%8%7%You build it,you own it andmanage performanceYou

47、 build it,but performance ismanaged by DevOps and SREswho then communicate back withthe developersYou build it,but someone elsemanages the performanceNone of the aboveEngineering and development culture within respondents organisations193.Development culture and KPIs Q3 2022|INFRASTRUCTURE VS APPLIC

48、ATIONS REPORT FOR SENTRYSRE/DevOps practitioners would prefer departments to have more autonomy and control over their tooling decisions rather than budgets being centralised%of professional SRE/DevOps practitioners(Q3 2022 n=141)29%23%22%17%9%28%30%23%17%2%Departments have abudgetand choose their o

49、wn tools,but there isa centralisedbudget for purchases above acertain limitDepartments have their ownbudget and they choose theirown toolsBudget and buying power iscentralised-Teams/departments do nothave purchasing autonomyTheres a centralised budgetbut departments/team havethe autonomy to choose t

50、heirown toolsNone of the aboveBudget and purchasing control within organisationCurrent setupPreferred setup203.Development culture and KPIs Q3 2022|INFRASTRUCTURE VS APPLICATIONS REPORT FOR SENTRYIn many modern software organisations,the“evolved SRE/DevOps team”where developers have ownership of app

51、lications and their performance is already in place.One contributing factor to this is ensuring that teams have the right tools and key performance indicators(KPIs)to effectively measure the health/performance of software applications in production.We asked respondents to report the KPIs they use fo

52、r such purposes within their organisation.Service availability,error/crash rate and system response time are the most commonly used KPIs.In addition to being used as reporting tools,these metrics can also be used proactively to estimate the probability that the system will perform as expected in fut

53、ure.Reactive metrics,which measure the mean time between the occurrence and detection(MTTD),identification of cause(MTTI)and repair(MTTR)of failures,are used less commonly.This emphasises that many SRE/DevOps practitioners are more focused on developing a proactive culture to bring stability assuran

54、ce to systems and operations.213.Development culture and KPIs Q3 2022|INFRASTRUCTURE VS APPLICATIONS REPORT FOR SENTRYFewer respondents use reactive metrics(e.g.mean-time-to-detect failures)as KPIs than proactive metrics(e.g.service availability)%of professional SRE/DevOps practitioners(Q3 2022 n=14

55、1)7%6%13%13%13%16%16%17%19%24%26%26%27%30%32%35%40%42%45%52%I dont use KPIs to measure system healthSaturationMean-time-to-detect(MTTD)failuresMean-time-to-identify(MTTI)cause of failuresChange failure rateLead time for changes to reach productionMean time between failuresRevenueMean-time-to-repair(

56、MTTR)failuresDeployment frequencyResource utilisationSystem traffic(demand)Service level objectives(SLOs)User engagementUser satisfactionService level agreements(SLAs)Metrics built into observability/monitoring toolsSystem response timeError/crash rateService availabilityKPIs used to measure health/

57、performance of a software application4.The cost of application issues/infrastructure outages234.The cost of application issues/infrastructure outagesQ3 2022|INFRASTRUCTURE VS APPLICATIONS REPORT FOR SENTRYIn this final chapter,we explore the impact of application issues and infrastructure outages in

58、 terms of the types of costs and damages organisations incur.We also provide data on the average loss per year in terms of productivity and financial costs incurred.Loss of developer productivity is the most likely outcome as a result of both application issues(55%)and infrastructure outages(54%).Al

59、most half of the SRE/DevOps practitioners indicated that these issues also result in more time spent on customer service(45%as a result of application issues and 42%as a result of infrastructure outages),as well as a significant impact on the productivity of SRE/DevOps teams(42%as a result of applic

60、ation issues and 44%as a result of infrastructure outages).The likelihood of such impacts demonstrates the importance of enabling SRE/DevOps teams to have sufficient time and energy to devote to proactive work to mitigate the risks of such issues occurring.In terms of productivity loss,30%of respond

61、ents estimate that their SRE/DevOps teams lose at least one person-month per year due to application issues-a considerable amount of time and effort.38%of SRE/DevOps practitioners estimate that their organisation spends$100K or more on additional developer efforts to resolve application issues,and 2

62、4%estimate that a similar amount is spent on SRE/DevOps time and effort.12%of respondents estimate that their organisation has had to pay out$100K or more in compensation in the past year for service-level agreement violations,and 12%also estimate that a similar amount has been spent on legal compli

63、ance fees and fines as a result of application issues.244.The cost of application issues/infrastructure outagesQ3 2022|INFRASTRUCTURE VS APPLICATIONS REPORT FOR SENTRYOrganisations experience a slightly wider variety of costs/damages as a result of application issues in comparison to infrastructure

64、outages%of professional SRE/DevOps practitioners(Q3 2022 n=139)55%45%42%38%28%25%22%21%15%12%13%54%42%44%28%25%21%15%10%10%11%16%Loss of developer productivityIncrease in customer service timeLoss of SRE/Dev Ops productivityDecrease in customer experience ratingsLoss in revenueDecrease in company br

65、and perceptionIncrease in user churnIncrease in security threats/vulnerabilities which can lead to breachesand resulting fines/compensation payoutsCompensation to customers for SLA violationsLoss in job securityWe have not experienced any costs or damagesWhat costs or damages has your organisation e

66、xperienced as a result of application issues or infrastructure outages affecting software in production?As a result of application issuesAs a result of infrastructure outages254.The cost of application issues/infrastructure outagesQ3 2022|INFRASTRUCTURE VS APPLICATIONS REPORT FOR SENTRY30%of respond

67、ents estimate that their SRE/DevOps teams lose more than one person-month in productivity per year fixing application issues%of professional SRE/DevOps practitioners(Q3 2022 n=117)1%5%25%40%25%5%None at allLess than one person-day1-7 person-days(betweenone day and a week)8-30 person-days(betweena we

68、ek and a month)31-365 person-days(between one month and ayear)More than 365 person-days(more than a year)On average,how much time would you estimate your SRE/DevOps teams spend per year fixing application issues?264.The cost of application issues/infrastructure outagesQ3 2022|INFRASTRUCTURE VS APPLI

69、CATIONS REPORT FOR SENTRY12%of respondents estimate that their organisation has had to pay out$100K or more in compensation in the past year for service-level agreement violations%of professional SRE/DevOps practitioners(Q3 2022 n=64)70%64%33%26%24%18%24%40%50%38%9%12%27%24%38%3%Compensation to cust

70、omers for SLA violationsLegal compliance/finesCustomer service effortDevOps resolution effortApplication developer resolution effortOver the past year,how much money(in$USD)would you estimate your SRE/DevOps team has spent on each of the following in order to fix application issues?$0-$10K$10K-$100K

71、$100K-$1MMore than$1MAppendix28AppendixQ3 2022|INFRASTRUCTURE VS APPLICATIONS REPORT FOR SENTRYSurvey respondents involvement in SRE/DevOps activities%of professional SRE/DevOps practitioners(Q3 2022 n=141)62%60%58%55%55%48%38%35%I approve codedeployments toproductionI monitor softwareand infrastruc

72、tureperformanceI use continuousintegration toautomatically buildand test my codechangesI use continuousdelivery/deploymentto automate mycode deploymentsI manage/buildCI/CD pipelinesI am a first responderto incidents/outagesI create automatedregression testsand/or validationchecksI programmaticallypr

73、ovision andmanage ITinfrastructures29SlashDataSlashData Ltd.Ltd.19-21 Hatton Gardens London,EC1N 8BAUnited Kingdom+44 845 003 8742 helloslashdata.coWe help you understand Developers.If you could speak to 30,000+developers what would you ask them?Simon JonesData Storytellersimon.jonesslashdata.coKonstantinos KorakitisDirector of Researchkonstantinosslashdata.co