Improve primary keys of clickhouse tables

Some of the DB tables could have their PRIMARY KEYs changed to improve query performance.

For example the PRIMARY KEY of the obs_web table is this:

ENGINE = ReplicatedReplacingMergeTree('/clickhouse/{cluster}/tables/ooni/obs_web_repl/{shard}', '{replica}')
PARTITION BY concat(substring(bucket_date, 1, 4), substring(bucket_date, 6, 2))
PRIMARY KEY (measurement_uid, observation_idx)

while the analysis table is this:

ENGINE = ReplicatedReplacingMergeTree('/clickhouse/{cluster}/tables/ooni/analysis_web_measurement/{shard}', '{replica}')
PARTITION BY substring(measurement_uid, 1, 6)
PRIMARY KEY measurement_uid
ORDER BY (measurement_uid, measurement_start_time, probe_cc, probe_asn, domain)

What this means is that it's triggering very large scans (effectively a full scan), when performing very common filtering operations over the measurement_start_time, probe_cc, probe_asn and domain.

Unfortunately it's not possible to change these with an ALTER query, so instead we are going to have to create the new table and copy the data over.

Start by locating the ClickHouse definitions for the obs_web and analysis_web_measurement tables shown in the issue. Review their current PRIMARY KEY and ORDER BY choices, then determine replacement keys for the listed filters and plan the new-table data copy. Done means the replacement tables are populated and common measurement_start_time, probe_cc, probe_asn, and domain queries no longer require effectively full scans.