Merge pull request #650 from etschannen/release-5.2

Enough concurrent readers could bring the cluster down when the cluster is rebooted
This commit is contained in:
Steve Atherton 2018-07-31 17:12:50 -07:00 committed by GitHub
commit 2d7922d550
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
3 changed files with 6 additions and 2 deletions

View File

@ -10,6 +10,11 @@ Bindings
* Java: ``FDBDatabase::run`` and ``FDBDatabase::read`` now use the ``Executor`` provided for executing asynchronous callbacks instead of the default one for the database. `(Issue #640) <https://github.com/apple/foundationdb/issues/640>`_
Fixes
-----
* A large number of concurrent read attempts could bring the database down after a cluster reboot. `(PR #650) <https://github.com/apple/foundationdb/pull/650>`_
5.2.7
=====

View File

@ -56,7 +56,7 @@ struct MasterProxyInterface {
getConsistentReadVersion.getEndpoint(TaskProxyGetConsistentReadVersion);
getRawCommittedVersion.getEndpoint(TaskProxyGetRawCommittedVersion);
commit.getEndpoint(TaskProxyCommit);
getKeyServersLocations.getEndpoint(TaskProxyGetKeyServersLocations);
//getKeyServersLocations.getEndpoint(TaskProxyGetKeyServersLocations); //do not increase the priority of these requests, because clients cans bring down the cluster with too many of these messages.
}
};

View File

@ -53,7 +53,6 @@ enum {
TaskProxyCommit = 8540,
TaskTLogConfirmRunningReply = 8530,
TaskTLogConfirmRunning = 8520,
TaskProxyGetKeyServersLocations = 8515,
TaskProxyGRVTimer = 8510,
TaskProxyGetConsistentReadVersion = 8500,
TaskDefaultPromiseEndpoint = 8000,