GitProtect has printed analysis displaying that incidents throughout main DevOps platforms rose 21% in 2025, whereas whole disruption time climbed to 9,255 hours.
The research recorded 607 incidents in 2025, up from 502 a yr earlier, and located that cumulative service disruption almost doubled from 4,755 hours. It additionally recognized 156 essential or main incidents, with a mixed period of 1,769 hours and 43 minutes, a 69% year-on-year improve in probably the most critical class.
The findings level to rising operational pressure throughout instruments utilized by software program improvement and IT groups. Instability was most pronounced within the second quarter, when 47 incidents have been logged, whereas June was probably the most risky month, with 21 incidents.
Outage patterns
Degraded efficiency accounted for 62% of all recorded incidents, or 374 circumstances, making it the commonest downside within the knowledge. Even so, these occasions made up 34% of whole downtime, or 3,059 hours, suggesting many have been widespread however shorter than different varieties of incidents.
Upkeep made up solely 4% of incidents however accounted for 30% of whole disruption time, making it the largest single driver of platform unavailability by period regardless of its comparatively low frequency.
Important and main incidents represented 26% of all incidents and 22% of whole disruption time. Minor points with no direct operational impression accounted for 8% of incidents and 16% of whole period.
Platform breakdown
Among the many platforms analysed, GitLab recorded the very best variety of essential and main incidents in 2025, with 62 incidents and greater than 754 hours of impression. Jira adopted with 44 incidents and almost 728 hours of downtime.
GitHub noticed 35 essential and main incidents with 172 hours of impression. Bitbucket recorded 15 such incidents, totalling 113 hours and 59 minutes.
The report additionally highlighted sharp month-to-month swings in disruption ranges throughout particular person platforms. For GitLab, July was the worst month, with 11 incidents inflicting greater than 252 hours of disruption, together with a virtually 50-hour outage linked to incorrectly deleted OAuth refresh tokens.
GitHub recorded its highest month-to-month incident rely in December, with 5 essential and main incidents. Its most extreme month-to-month impression got here in March, when two incidents prompted greater than 58 hours of degraded efficiency, together with a chronic outage triggered by expired inside credentials.
Bitbucket’s most disruptive month was Might, with 5 incidents producing greater than 84 hours of impression, together with a 49-hour disruption affecting pipeline execution. In Jira, June introduced 9 incidents and virtually 48 hours of disruption. The longest single incident that month lasted virtually 24 hours and affected Jira Software program and Jira Service Administration due to Forge-related errors within the Singapore area.
Operational price
GitProtect connected an estimate of misplaced engineering productiveness to the cumulative outage time. Utilizing a median developer fee of USD $80 an hour, it estimated that the overall disruption equated to greater than USD $740,000 in misplaced productiveness alone.
That calculation doesn’t embrace wider industrial results equivalent to delayed releases, service credit, buyer help prices or the impression on inside IT operations. The figures nonetheless underline how disruption in developer instruments can immediately have an effect on engineering groups that depend on cloud-based companies for code internet hosting, collaboration, deployment and ticketing.
The analysis lined safety breaches, outages and efficiency degradation throughout main improvement platforms utilized by thousands and thousands of individuals. Its figures counsel rising stress on cloud-based software program improvement infrastructure, not solely within the variety of incidents but in addition in how lengthy groups stay affected when issues happen.
The info means that even extensively used platforms with established resilience measures stay uncovered to service interruptions. It additionally signifies that rising downtime, moderately than incident rely alone, is the clearest signal of a altering danger profile for improvement groups.
Upkeep home windows, degraded efficiency and extreme incidents every contributed otherwise to the general disruption image, with deliberate and unplanned upkeep accounting for a bigger share of whole downtime than the variety of occasions alone would possibly counsel. The sample reveals that frequency by itself doesn’t seize the operational burden positioned on customers.
The evaluation concludes that the hole between incident quantity and incident impression has widened, as longer disruptions account for a rising share of general downtime throughout the DevOps software program market.