at qserv level: if query cost estimate does not catch it, then abort when limit exceeded
at webserv level: if result can't be processed e.g. because running out of memory, return part of the result that can be returned, and include metadata explaining what happened
need to understand how to abort query that is in the middle of streaming - how to force it to abort reasonably
create two new epics for Qserv:
streaming results. (Created: )
detecting and handling unexpected conditions during query execution (e.g. bad chunk, hit per user resource limit). (Created: )
webserv client
in python
for testing webserv
can be used as a reference implementation for future clients (e.g. java based)
naming modules
concern that "webserv", "webserv_client" are too generic
imgserv is a bit like qserv. Should be for code related to retrieving and serving data over http. Any other code that is generally useful and involves algorithms should be moved to afw