问题
I've been trying explore if there's a way to retry the createTask function. The reason being is because I keep on encountering deadline exceeded error from time to time like so:
Error: 4 DEADLINE_EXCEEDED: Deadline Exceeded at Object.exports.createStatusError (/srv/node_modules/grpc/src/common.js:91:15) at Object.onReceiveStatus (/srv/node_modules/grpc/src/client_interceptors.js:1204:28) at InterceptingListener._callNext (/srv/node_modules/grpc/src/client_interceptors.js:568:42) at InterceptingListener.onReceiveStatus (/srv/node_modules/grpc/src/client_interceptors.js:618:8) at callback (/srv/node_modules/grpc/src/client_interceptors.js:845:24)
Reading on the code behind the createTask function, I found out that the default timeout configuration was just 10 seconds.
At the moment, I have tried to extend the timeout to 30s (which I believe is the maximum) by doing this:
try {
console.log("Sending task %j", task);
const callOptions = {
timeout: 30000
};
// Send create task request.
const [response] = await client.createTask(request, callOptions);
const name = response.name;
console.log(`Created task ${name}`);
} catch (error) {
console.error("CREATE_TASK_ERROR::", error);
}
And it appears that it works. However, I would also like to cover the case if the API wasn't able to respond within 30 seconds.
I've tried this code:
try {
console.log("Sending task %j", task);
const callOptions = {
timeout: 2000, // I've set it to 2 seconds to be able to reproduce the deadline exceeded error easily
retry: {
initial_retry_delay_millis: 100,
retry_delay_multiplier: 1.3,
max_retry_delay_millis: 60000,
initial_rpc_timeout_millis: 20000,
rpc_timeout_multiplier: 1.0,
max_rpc_timeout_millis: 20000,
total_timeout_millis: 300000
}
};
// Send create task request.
const [response] = await client.createTask(request, callOptions);
const name = response.name;
console.log(`Created task ${name}`);
} catch (error) {
console.error("CREATE_TASK_ERROR::", error);
}
But I don't see the createTask being retried. But based on the comment here, we should be able to override the default settings including retries.
What am I doing wrong? Please help.
回答1:
It seems to that callOptions is wrong.
const callOptions = {
timeout: 2000, // I've set it to 2 seconds to be able to reproduce the deadline exceeded error easily
retry: {
backoffSettings: {
initialRetryDelayMillis: 100,
retryDelayMultiplier: 1.3,
maxRetryDelayMillis: 60000,
initialRpcTimeoutMillis: 20000,
// rpc_timeout_multiplier: 1.0, not exists
maxRpcTimeoutMillis: 20000,
totalTimeoutMillis: 300000
}
}
};
See:
- https://googleapis.github.io/gax-nodejs/global.html#CallOptions
- https://googleapis.github.io/gax-nodejs/global.html#RetryOptions
- https://googleapis.github.io/gax-nodejs/global.html#BackoffSettings
But I think that to set retry parameters using cli is better.
See:
- https://cloud.google.com/tasks/docs/creating-appengine-queues#retry
来源:https://stackoverflow.com/questions/55577504/how-to-override-the-retry-configuration-for-google-cloud-tasks-node-js-client