Thread: MUSTANG web service update at 2:30 PM PDT

Hi Folks-

Sorry for the outage that occurred. There were some changes in the metrics that caused an issue with the database initialization. I had to pull a few wires and then let the system reset. I think things are now back in proper operation, but please let me know if you have issues.

On another note, we are testing some new metrics that make use of time series availability. Because of the enormous time series catalog we have, please don't call any of the ts_ metrics with large time windows or extensive wildcard expansions right now. There are some optimizations needed to cut down the memory usage in these new metrics so that the server doesn't hang. Feel free to try them out with small-ish queries, though. There will be more metrics to come after I get the mechanics optimized.

Thank you for allowing us to improve QA service to you. This system is tricky, and we'll continue to work on making it work better than ever.

-Rob

On Nov 2, 2017, at 1:39 PM, Robert Casey <rob<at>iris.washington.edu> wrote:


Hi MUSTANG users-

We are going to implement a rolling update of the MUSTANG web services at 2:30PM. There should be no disruption of service in the update and we will evaluate how it's working in production to assure that it's stable and functional.

Thanks for your patience,

-Rob