2cf5009ff5
Make lock2's retryInterval smaller so it grabs the lock as soon as lock1's renewer fails to renew in time. Fix the logic to test if lock1's leader channel gets closed: we don't need a goroutine, and the logic was broken in that if we timed out we'd never write to the blocking channel we then try to read from. Moreover the timeout was wrong. |
||
---|---|---|
.. | ||
alicloudoss | ||
azure | ||
cassandra | ||
cockroachdb | ||
consul | ||
couchdb | ||
dynamodb | ||
etcd | ||
foundationdb | ||
gcs | ||
manta | ||
mssql | ||
mysql | ||
postgresql | ||
s3 | ||
spanner | ||
swift | ||
zookeeper |