Skip to content

fix(grpc): Add keepalive and fix reconnect issue #777

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
+44 −24
Open

fix(grpc): Add keepalive and fix reconnect issue #777

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 14 additions & 0 deletions ...niffer/apm-agent-core/src/main/java/org/apache/skywalking/apm/agent/core/conf/Config.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -210,6 +210,20 @@ public static class Collector {
* How long grpc client will timeout in sending data to upstream.
*/
public static int GRPC_UPSTREAM_TIMEOUT = 30;
/**
* The interval in seconds to send a keepalive ping to the backend.
* If this is less than or equal to 0, the keepalive is disabled.
*
* This maps to `collector.grpc_keepalive_time` in agent.config.
*/
public static long GRPC_KEEPALIVE_TIME = 60L;
/**
* The timeout in seconds to wait for a keepalive ack from the backend.
* If the ack is not received within this time, the connection is considered dead.
*
* This maps to `collector.grpc_keepalive_timeout` in agent.config.
*/
public static long GRPC_KEEPALIVE_TIMEOUT = 30L;
/**
* Get profile task list interval
*/
Expand Down
7 changes: 7 additions & 0 deletions ...apm-agent-core/src/main/java/org/apache/skywalking/apm/agent/core/remote/GRPCChannel.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -27,6 +27,8 @@
import io.grpc.netty.NettyChannelBuilder;
import java.util.LinkedList;
import java.util.List;
import java.util.concurrent.TimeUnit;
import org.apache.skywalking.apm.agent.core.conf.Config;

public class GRPCChannel {
/**
Expand All @@ -39,6 +41,11 @@ private GRPCChannel(String host, int port, List<ChannelBuilder> channelBuilders,
List<ChannelDecorator> decorators) throws Exception {
ManagedChannelBuilder channelBuilder = NettyChannelBuilder.forAddress(host, port);

if (Config.Collector.GRPC_KEEPALIVE_TIME > 0) {
channelBuilder.keepAliveTime(Config.Collector.GRPC_KEEPALIVE_TIME, TimeUnit.SECONDS)
.keepAliveTimeout(Config.Collector.GRPC_KEEPALIVE_TIMEOUT, TimeUnit.SECONDS);
}

NameResolverRegistry.getDefaultRegistry().register(new DnsNameResolverProvider());

for (ChannelBuilder builder : channelBuilders) {
Expand Down
41 changes: 17 additions & 24 deletions ...nt-core/src/main/java/org/apache/skywalking/apm/agent/core/remote/GRPCChannelManager.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -130,34 +130,27 @@ public void run() {
String server = "";
try {
int index = Math.abs(random.nextInt()) % grpcServers.size();
if (index != selectedIdx) {
selectedIdx = index;
selectedIdx = index;

server = grpcServers.get(index);
String[] ipAndPort = server.split(":");
server = grpcServers.get(index);
String[] ipAndPort = server.split(":");

if (managedChannel != null) {
managedChannel.shutdownNow();
}

managedChannel = GRPCChannel.newBuilder(ipAndPort[0], Integer.parseInt(ipAndPort[1]))
.addManagedChannelBuilder(new StandardChannelBuilder())
.addManagedChannelBuilder(new TLSChannelBuilder())
.addChannelDecorator(new AgentIDDecorator())
.addChannelDecorator(new AuthenticationDecorator())
.build();
reconnectCount = 0;
reconnect = false;
notify(GRPCChannelStatus.CONNECTED);
} else if (managedChannel.isConnected(++reconnectCount > Config.Agent.FORCE_RECONNECTION_PERIOD)) {
// Reconnect to the same server is automatically done by GRPC,
// therefore we are responsible to check the connectivity and
// set the state and notify listeners
reconnectCount = 0;
reconnect = false;
notify(GRPCChannelStatus.CONNECTED);
LOGGER.debug("Attempting to reconnect to gRPC server {}. Shutting down existing channel if any.", server);
if (managedChannel != null) {
managedChannel.shutdownNow();
}

managedChannel = GRPCChannel.newBuilder(ipAndPort[0], Integer.parseInt(ipAndPort[1]))
.addManagedChannelBuilder(new StandardChannelBuilder())
.addManagedChannelBuilder(new TLSChannelBuilder())
.addChannelDecorator(new AgentIDDecorator())
.addChannelDecorator(new AuthenticationDecorator())
.build();
LOGGER.debug("Successfully reconnected to gRPC server {}.", server);
reconnectCount = 0;
reconnect = false;
notify(GRPCChannelStatus.CONNECTED);

return;
} catch (Throwable t) {
LOGGER.error(t, "Create channel to {} fail.", server);
Expand Down
6 changes: 6 additions & 0 deletions apm-sniffer/config/agent.config
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -100,6 +100,12 @@ collector.properties_report_period_factor=${SW_AGENT_COLLECTOR_PROPERTIES_REPORT
collector.backend_service=${SW_AGENT_COLLECTOR_BACKEND_SERVICES:127.0.0.1:11800}
# How long grpc client will timeout in sending data to upstream. Unit is second.
collector.grpc_upstream_timeout=${SW_AGENT_COLLECTOR_GRPC_UPSTREAM_TIMEOUT:30}
# The interval in seconds to send a keepalive ping to the backend.
# If this is less than or equal to 0, the keepalive is disabled.
#collector.grpc_keepalive_time=${SW_AGENT_COLLECTOR_GRPC_KEEPALIVE_TIME:60}
# The timeout in seconds to wait for a keepalive ack from the backend.
# If the ack is not received within this time, the connection is considered dead.
#collector.grpc_keepalive_timeout=${SW_AGENT_COLLECTOR_GRPC_KEEPALIVE_TIMEOUT:30}
# Sniffer get profile task list interval.
collector.get_profile_task_interval=${SW_AGENT_COLLECTOR_GET_PROFILE_TASK_INTERVAL:20}
# Sniffer get agent dynamic config interval.
Expand Down