Hawkin
/
arangodb
mirror of https://gitee.com/bigwinds/arangodb
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

Fix shutdown deadlock regarding comm tasks (#9204) 

* Wait for _commTasks in unprepare, that is after Cluster::stop

* Chose better method names

* Revert "Chose better method names"

This reverts commit 91e821348740c655f47207af7e570075f2241895.

* Revert "Wait for _commTasks in unprepare, that is after Cluster::stop"

This reverts commit 6551ae90d74fc046369fdb97cc5872706ce1a184.

* Next try, stop ClusterComm threads earlier
This commit is contained in:
Tobias Gödderz 2019-06-07 13:23:33 +02:00 committed by Michael Hackstein
parent 6b16d1c0a5
commit b632d58c80
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
arangod/Cluster/ClusterComm.cpp
View File

@ -1150,7 +1150,7 @@ void ClusterCommThread::run() {
TRI_ASSERT(_communicator != nullptr);
LOG_TOPIC("74eda", DEBUG, Logger::CLUSTER) << "starting ClusterComm thread";
auto lastAbortCheck = std::chrono::steady_clock::now();
while (!isStopping()) {
while (!application_features::ApplicationServer::isStopping()) {
try {
if (std::chrono::steady_clock::now() - lastAbortCheck >
std::chrono::duration<double>(3.0)) {