Feels like we need to get this working for the all-hands on Thursday.

IIRC, those scripts are not maintained by us.

But I had a look at them, and they produce some empty files, and show errors
to connect to s1 slave.

So the issue sounds related to db1047 being unusable (RT: 7385).

(In reply to christian from comment #2)

So the issue sounds related to db1047 being unusable (RT: 7385).

db1047 is back and usable.

Now the script yields different error messages, like

Generating Successful Main namespace edits (per day)
Issue executing SQL for successful-edits-main ('NoneType' object is not iterable)
Generating Uploads failed because of an upload error (per day)
Issue executing SQL for error-uploads ('NoneType' object is not iterable)

when running the SQL (for example for error-uploads) by hand, the
database responds with messages like

ERROR 1194 (HY000): Table 'MobileWebUploads_5383883' is marked as crashed and should be repaired

which is due to complications around EventLogging's database
migration. I guess we cannot do much until the database migration
finished.

(Not sure if the maintainers of the scripts that generate mobile-reportcard
have backup data somewhere, or can pin some csvs to a previous known good
file. You'd have to clarify with them.)

EventLogging database is queryable again, and the mobile-reportcard
came back.

But when looking at the data in the tables of EventLogging, it seems
replication still has to catch up.

EventLogging tables caught up.
Graphs seem ok again.