Apr 14, 2022 16:02:00

It took 9 days for Atlassian to officially announce the outbreak of the failure, and it took 9 days to get the user's anger.

Atlassian, which develops many products such as the project management tool 'Jira Software', the task management tool 'Trello', and the workspace tool 'Confluence', has

partially stopped services. At the beginning of the failure, about 400 companies were affected, and after that, although the scope of impact has narrowed due to the restoration of functions, as of April 14, 2022, half of the users are still unable to use the service. .. Regarding this matter, tech expert Gergely Orosz points out the accumulated complaints that 'Atlassian lacks communication with users'.

The Scoop: Inside the Longest Atlassian Outage of All Time
https://newsletter.pragmaticengineer.com/p/scoop-atlassian

Engineering mistakes happen. The most inexcusable thing is not communicating wit ... | Hacker News
https://news.ycombinator.com/item?id=31020562

Atlassian Service Suspension Affecting Some Customers | Atlassian Japan Official Blog | Atlassian Co., Ltd.
https://www.atlassian.com/ja/blog/april-2022-outage-update

According to Atlassian's announcement, the service was suspended at Coordinated Universal Time around 20:12 on Monday, April 4, 2022 (5:12 on April 5, Japan time). The two causes are 'mistake of the ID of the standalone application that I originally wanted to disable and the ID of the entire cloud site' and 'I executed it with' permanent deletion 'instead of' recoverable deletion 'when executing the deletion script'. As a result, about 400 sites have been deleted.

The following pages keep track of what Atlassian was reporting about this down and its response. As of 11:00 on April 5, two hours after the start of the investigation, there is a report that 'the cause has been found and mitigation measures are being planned', but there is no mention of the specific cause or work.

Jira Software Status --Multiple sites showing down / under maintenance
https://jira-software.status.atlassian.com/incidents/gms0m979jxbr

After all, the detailed report was made on April 12th, with CTO Sri Viswanath updating the engineer blog.

April 2022 outage update
https://www.atlassian.com/engineering/april-2022-outage-update

Atlassian often claims that he is 'in direct communication with users' about the matter, but Orosz has received comments from multiple users that 'there was no individual contact.' 'All customer complaints are due to Atlassian's lack of communication,' Orosz said.

He also points out that the impact of this case will have a negative impact on long-term cloud sales in the form of 'loss of reputation' rather than 'loss of revenue.' Uber has already stopped using HipChat because of a security problem with the Atlassian communication tool 'HipChat' twice in 2015 and 2017.

HipChat Security Notice and Password Reset | Atlassian Japan Official Blog | Atlassian Co., Ltd.
https://www.atlassian.com/ja/blog/hipchat-security-notice-and-password-reset

In addition, Atlassian is moving away from server products and moving to cloud services, but this may force us to extend the support period for our server products. Orosz says Atlassian is a good option if you don't want to lose big customers.

In addition, Mr. Orosz accused Atlassian of this response as 'lacking the agility that a high-tech company should be', and said that internal reforms and investments will be made to improve problem handling in the future. I'm closing when I expect it.