pm2 commands can't work

How:

I has an app(host by pm2) that needs 25%-30%(400mb--600mb) memory for one cluster, and the app has 2 cluster. When I started a batch task ,the app will occupy 100% cpu for a while. and then the pm2 command can't work. my app is still responsible

Result:

1 . The pm2 commands(pm2 log,pm2 list) can't work , with a response like this:

node[503684]: ../src/node_platform.cc:61:std::unique_ptr<long unsigned int> node::WorkerThreadsTaskRunner::DelayedTaskScheduler::Start(): Assertion `(0) == (uv_thread_create(t.get(), start_thread, this))' failed.
 1: 0xb0a860 node::Abort() [node]
 2: 0xb0a8de  [node]
 3: 0xb7a11e  [node]
 4: 0xb7a1e6 node::NodePlatform::NodePlatform(int, v8::TracingController*) [node]
 5: 0xacd05e node::InitializeOncePerProcess(int, char**, node::InitializationSettingsFlags, node::ProcessFlags::Flags) [node]
 6: 0xace3a9 node::Start(int, char**) [node]
 7: 0x7f5ceef64ac5 __libc_start_main [/lib64/libc.so.6]
 8: 0xa4067c  [node]
Aborted

2. the apps that started by pm2 works ok and the pm2 process is still in my process list

Reproduce the failure by running pm2 log or pm2 list while the two-cluster application is under high CPU usage. Start with the reported Node.js WorkerThreadsTaskRunner assertion and compare the failed commands with the still-running PM2-managed processes; done means the commands remain usable under the described load.