fix(cassandra) ensure single coordinator in migrations #2326
Conversation
This ensures we respect the proper 'single coordinator' pattern for migrations when one of them is using the DAO's `find_all()` method. * ensure `find_all()` only uses the migrations coordinator. * ensure we wait for schema consensus before doing so, since the coordinator will have to get responses about the table's content from its peers. * perf: only wait for schema consensus if we ran some migrations at all.
Implement a new `cassandra_schema_consensus_timeout` property to increase the C* `max_schema_consensus_wait` value. Particularly useful for clusters where the inter-nodes communication seems to be slow and the schema changes during migrations can take more than the default of 10s, and make the migration fail unnecessarily.
kong/dao/db/cassandra.lua
Outdated
| -- before performing such a DML query | ||
| local ok, err = self:wait_for_schema_consensus() | ||
| if not ok then | ||
| return nil, "could not wait for schema consensus: " .. err |
There was a problem hiding this comment.
"failed waiting for schema consensus"
kong/dao/factory.lua
Outdated
| local ok, err = self.db:wait_for_schema_consensus() | ||
| if not ok then | ||
| return ret_error_string(self.db.name, nil, | ||
| "could not wait for schema consensus: " .. err) |
There was a problem hiding this comment.
both forms are used intermittently in the codebase
There was a problem hiding this comment.
I'd say that only impatient people 'cannot wait' 😄
|
The "do not merge" label is for refused PRs.
Sadly not, or else it would be included. |
Actually the driver has little to do with this all. |
surely we close those? this label at least is more descriptive than a foot note in the original post. |
It is sometimes more complicated than that.
Sounds just like the definition of a timeout! |
This ensures we update the Nginx time between schema consensus timeout checks and also adds the ability to manually add and remove C* peers. thibaultcha/lua-cassandra@1.1.1...1.2.1
thibaultcha
force-pushed
the
fix/single-coordinator-migrations
branch
from
9ec9e79 to
ddeea29
Compare
Updated them in the end 😉 |
Summary
This ensures we respect the proper 'single coordinator' pattern for
migrations when one of them is using the DAO's
find_all()method.Full changelog
find_all()only uses the migrations coordinator.coordinator will have to get responses about the table's content from
its peers.
driver.
cassandra_schema_consensus_timeoutproperty.when testing for a schema consensus timeout.
NOTE: Do NOT merge yet.