Hi Justin,

We are in Enterprise subscription and our users use JSM for 24/7 support, and the JSM instance is highly dependent on assets. Assets not available means no one can create tickets. I got to know from Atlassian support these activities happen during:

• Asia Pacific and Central Europe - Mon, Wed, Thurs, 9am AEST / 10am AEDT,

• Americas, Western Europe - Mon, Wed, Thurs, 1pm AEST / 2pm AEDT

It would be great if Atlassian take necessary actions to these deployments causes zero downtime.

Best Regards,

Suraj

Is this related to or the cause of this issue linked below, which occurs daily from 6:45pm-10:45pm UTC?  I'd like to identify which one to follow as the parent issue and bug resolution.  Thanks.

JSDCLOUD-14308 JSM Assets API latency causing failures in dependent operations - Create and track feature requests for Atlassian products.

7260236ab89d  - since there is no workaround - can we please be notified when assets will have a brief downtime AHEAD Of time?  

This has an extremely negative impact for our customers .

It would be helpful to receive advance notice of such outages or to see an update on the status page indicating outages. Will Atlassian provide this information in the future?

Moving the deployments "out of business hours where possible" does not help us at all because we have orchestration jobs that run every hour so any downtime will negatively impact us.

We can sometimes go for weeks with no failures in our scheduled orchestration jobs but at other times (like that past week or so) we are getting hit by this several times per week (four times in the last five days in fact). Since it is acknowledged as a bug I don't understand why it is still "gathering interest", I would expect the team to figure out how to do non-disruptive deployments ASAP.

When we first encountered this bug last year I modified our orchestration to first call the list object schema API then wait 30 seconds since I was advised by Support that this would work around the bug but that has not been the case lately, we end up having to rerun the job later. It is also annoying that the outages are not in roughly the same time window every day, sometimes they are very early morning PST and sometimes a few hours before midnight.

Once more, I am puzzled by the fact that this is an issue. Assets are a Premium Feature and Enterprise-grade customers tend to use them (or they are even required to fulfill all their requirements). In my particular case, the support runs 24/7 and the JSM instance is highly dependent on assets. Whilst in "usual" user experience, this is an annoyance, with automations this is a no-go. The automation fails in the middle of the night, causing tickets to be stuck as the trigger passed, and the automation fails for no reason.

Hello Justin, 

Do you have any updates on when the planned downtime will occur for each region?

Best regards, 

Dimitris 

Hi Justin,

Thanks for update.  I wanted to mention though on point:

"Moving the regular downtime windows to be region specific"

Our company accepts global JSM portal requests so there is no downtime that can really work for us.  But hopefully minimizing downtime will help.  It would be preferred to have no downtime, or be able to re-trigger or re-try any Automation Rules somehow after a delay.

Thanks,

Sean

Hi all,

Is there any specific ETA for the planned downtime per region?

Until then, it would be beneficial for deployments to happen at times other than xx:00 and xx:30 hours, as these are the available options for scheduled automation.

Please, consider this as a temporary workaround!

Thank you,

Dimitris