As far as I can tell, the new Diagnostics system does not support integrations (at least our Discord integration that was working in deprecated Cloud Diagnostics). Our team found this extremely useful in the past as we would be notified immediately and could tell if there was a major problem as opposed to having to remember to manually check it and potentially missing something for long periods of time.
In addition, the previous crash reporting system seemed to be almost instant. If we saw an error occur, we could immediately refresh and see the issue. The current system seems to take several minutes to populate new reports. This was extremely useful when we had streamers playing our game because we would be actively watching, get the report, and have immediate context as to what the user was doing and what went wrong. Now we have to try to remember what someone was doing for the past 10 minutes or so.
There also doesn’t seem to be a way to clear issues which was useful before. I could know which issues I already fixed by tagging them as closed. At least a way to tag issues would be helpful and save time not looking at reports we’ve already fixed and having to figure out if they are still relevant or not.
Thank you for that feedback. We really appreciate these types of insights.
Regarding the first point. You had a custom integration that would message Discord with status messages? I would love to hear more about this integration so I can pass this feedback on to the team.
I’ve asked someone from our team to explore potential workarounds for your other points. That feedback will also be shared with the Diagnostics team.
Which, the point being is that this already exists on the dashboard for Unity, just not for the new Diagnostics for some reason, unless I am missing it somewhere.
To provide you with an update to this feedback/thread:
Integrations are coming soon for Engine Diagnostics; we are in the final phases of development to reach parity with Cloud Diagnostics.
There is currently no way to add tags to crash report instances, and it is not a planned feature due to it not being heavily used with Cloud Diagnostics; however, I have shared your feedback as a +1 to have this feature implemented.
Regarding the delay in reports arriving on the dashboard, that is a known limitation of the new backend pipeline. During my own testing, I observed a 3-to 4-minute delay in reports being presented on the dashboard. This is noted as a concern, in particular with the use case you mentioned here, and effectiveness and speed is always a priority for our services.
Hopefully, this helps although I realize it may not be ideal for the most part.
I’d say the system on the whole is an improvement, and of those 3, Integrations are the most critical in my opinion. Having something to notify you of a problem helped tremendously with our last project and it just means we don’t have to hover over the diagnostics for new problems to come in.
I can live with the other two limitations for sure, but just in the spirit of feedback:
Tags: I wouldn’t say that we need tags specifically, because I will also admit that I didn’t use tags for anything except for when I closed an issue I fixed so that I could keep track. We also had integrations that would create Jira tickets, so this still isn’t critical, but I think of these reported issues as tickets, and it’s just simpler to work in a ticketing system where you can close tickets.
Delay: This also isn’t really a huge deal. My case of watching streamers play my games and getting pings in the moment a problem happened where I could get context, or even clip it because it was that instantaneous, is nice but niche, and it’s not like I’m able to watch every single person ever who streams a game I make. I honestly wouldn’t have even mentioned it if I hadn’t experiences the very fast times with Cloud Diagnostics.
If I get the integrations and nothing else, that will still make solving problems way more efficient.
Thanks for the follow-up and the detailed feedback. It is appreciated!
I definitely agree with the importance of integrations
I understand the rest of the detailed feedback as well. The streamer use-case is quite interesting, as I haven’t seen that before. That’s pretty cool
This sounds good, and while I don’t have a specific ETA for integrations, the backend work is mostly completed now, so it shouldn’t be long before we see that feature.
We upgraded to 6000.3.10f1 last week, after which we removed the deprecated cloud diagnostics package from the project. While testing I noticed that the webhooks are not working anymore. I am assuming this thread is related to that.
I had integrated teams channel to the error notifications, after the upgrade and removal of cloud diagnostics the alert have stopped. I have checked the webhook urls they are working fine. I am assuming unity backend isnt triggering the webhooks.
Are you able to let us know the ETA at this point of time ?
I re-imported and used the deprecated cloud diagnostics for the teams integration, the webhooks are working as expected. We will continue using cloud diagnostics until the webhook feature support has been implemented to the new diagnostic framework
I agree, and have reported the lack of ability to close bugs to the team. However, they have deemed it an unnecessary feature that hardly anybody used in the previous version of diagnostics? I find this very hard to believe, but I would like to hear from others how they manage the error list in diagnostics.
Hello,
Just to bump this, it is potentially a big blocker for us for upgrading if webhooks aren’t working on new Diagnostics.
Can someone from Unity team confirm on whether this should work now or on any ETA for that?
