Influx Query with lots of select or better approach

Question

I am sorry if this quest sounds like something simple. But, I have just begun to work with Influx.

I have the following data set in my influx DB.

> select * from "productA" where "key" = 'auth-service' and time >= now() - 2d
name: productA
time                key                     value
----                ---                     -----
1606149940282823100 auth-service             200
1606150840121208900 auth-service             200
1606151739988156400 auth-service             200
1606152640068272500 auth-service             501
1606153540171206600 auth-service             200
1606154440452052500 auth-service             200
1606155340367436300 auth-service             501
1606156240083789100 auth-service             501
1606157140017362300 auth-service             200
1606158040164140600 auth-service             200
1606158940108503600 auth-service             200
1606159840290334300 auth-service             200
.
.
.
.
.
1606309240026904100 auth-service             200
1606310140215219500 auth-service             200
1606311040035353800 auth-service             200
1606311940135475600 auth-service             200
1606312840182629300 auth-service             404
1606313740375395900 auth-service             200
1606314640279048300 auth-service             200
1606315540205045900 auth-service             200
1606316440125233400 auth-service             200
1606317340177812500 auth-service             404
1606318240399528000 auth-service             200
1606319140427589200 auth-service             200
1606320039995366300 auth-service             200
1606320940496533400 auth-service             200
1606321840238763300 auth-service             200

I am trying to write a query that should give me results where the value is NOT 200 and the next entry where the value turns back to 200. So, if there is 5 entry in the sequence with value not equal to 200 then its should give me the first entry where the value was reported NOT 200 and this should be grouped by dates.

In simple word, this measurement data get stored using a script which checks the HTTP Response Code of auth service of a productA every 15 mins. I am trying to create a web page where users can see a list of events for the past 15 days where they can see when the auth-service failed and turned back to success with timestamps. If this event occurs multiple times then the page will show that as well.

for example

Date : 11-24-20

11:00 AM - productA Auth service Failed
12:15 PM - productA Auth Service Restored

5:00 PM - productA Auth service Failed
5:30 PM - productA Auth Service Restored

Date : 11-25-20

11:00 AM - productA Auth service Failed
12:15 PM - productA Auth Service Restored

So far my approach was to get the list of all entries where the status code is not 200 for the last 15 days and then use that data to make another query where I use the timestamps from the first query to get the next entry where the value is 200. But that approach doesn't handle multiple occurrences of such event in a single day.

Any optimal solution is appreciated!

Thanks,

Mike