[Ray Log] remove glog dependency #16077

qicosmos · 2021-05-26T07:52:54Z

Why are these changes needed?

We use spdlog for logging now, however we still rely on glog, because we need its GetCallTrace method, actually it can be replaced with absl.stacktrace, so this pr will remove glog dependency.

Related issue number

Checks

I've run scripts/format.sh to lint the changes in this PR.
I've included any doc changes needed for https://docs.ray.io/en/master/.
I've made sure the tests are passing. Note that there might be a few flaky tests, see the recent failures at https://flakey-tests.ray.io/
Testing Strategy
- Unit tests
- Release tests

kfstorm · 2021-05-26T08:57:06Z

@qicosmos Can you try to send SIGABRT to raylet and paste the stack trace logs with and without glog here?

kfstorm · 2021-05-26T09:00:55Z

We should delete this and all glog-related patches.

ray/bazel/ray_deps_setup.bzl

Lines 184 to 194 in 498da13

    
           auto_http_archive( 
        
               name = "com_github_google_glog", 
        
               url = "https://github.com/google/glog/archive/925858d9969d8ee22aabc3635af00a37891f4e25.tar.gz", 
        
               sha256 = "fb86eca661497ac6f9ce2a106782a30215801bb8a7c8724c6ec38af05a90acf3", 
        
               patches = [ 
        
                   "//thirdparty/patches:glog-log-pid-tid.patch", 
        
                   "//thirdparty/patches:glog-stack-trace.patch", 
        
                   "//thirdparty/patches:glog-suffix-log.patch", 
        
                   "//thirdparty/patches:glog-dump-stacktrack.patch", 
        
               ], 
        
           )

kfstorm · 2021-05-26T09:05:04Z

src/ray/util/logging.cc

+  local_stack.resize(50);
+  absl::GetStackTrace(local_stack.data(), 50, 1);
+  static constexpr size_t buf_size = 1 << 14;
+  std::unique_ptr<char[]> buf(new char[buf_size]);


Why do we need unique_ptr here?

kfstorm · 2021-05-26T09:05:48Z

src/ray/util/logging.cc

+  std::vector<void *> local_stack;
+  local_stack.resize(50);
+  absl::GetStackTrace(local_stack.data(), 50, 1);
+  static constexpr size_t buf_size = 1 << 14;


Suggested change

static constexpr size_t buf_size = 1 << 14;

static constexpr size_t buf_size = 16 * 1024;

kfstorm · 2021-05-26T09:07:34Z

src/ray/util/logging.cc

-  google::InitGoogleLogging(app_name_.c_str());
-  int level = GetMappedSeverity(static_cast<RayLogLevel>(severity_threshold_));
-#endif
+#ifdef RAY_USE_SPDLOG


Can we remove RAY_USE_SPDLOG as well?

#ifdef RAY_USE_SPDLOG typedef ray::SpdLogMessage LoggingProvider; #else typedef ray::CerrLog LoggingProvider; #endif
If we don't define RAY_USE_SPDLOG, ray::CerrLog will be used. If we remove RAY_USE_SPDLOG, how to provide a default logging provider?

I don't think CerrLog is being used. We can delete it as well.

I am okay with just deleting glog to make iteration faster btw. (and we can remove Cerrlog in the separate PR)

kfstorm · 2021-05-26T09:11:30Z

src/ray/util/logging.cc

-    return;
-  }
-  google::InstallFailureSignalHandler();
-  google::InstallFailureWriter(&WriteFailureMessage);


If we remove this, who will call WriteFailureMessage? I see that you removed all the signal handler-related code, but if we don't install the signal handler, how would we print stack trace on receiving a signal?

Maybe it should not be removed, i will test signal handler.

We need to rewrite singalhandler if these functions of glog should be removed.

Right, i will deal with signal handler later.

kfstorm · 2021-05-26T09:13:21Z

src/ray/util/logging.h

-/// have added a patch to allow glog to return the current call stack information
-/// through the internal interface. This function `GetCallTrace` is a wrapper
-/// providing a new detection function for debug or something like that.
+/// This function return the current call stack information.


Suggested change

/// This function return the current call stack information.

/// This function returns the current call stack information.

qicosmos · 2021-05-26T11:28:29Z

@qicosmos Can you try to send SIGABRT to raylet and paste the stack trace logs with and without glog here?

I will do a test.

…ve_glog_dependency

rkooo567 · 2021-05-29T01:11:18Z

Can you also post some of stacktrace example? (e.g., show me how stacktrace looks like when SIGBUS happens)

rkooo567 · 2021-06-01T02:01:24Z

@qicosmos can you post how the stacktrace looks like? Also does it make sense to write a python test to see if it prints logs on some signals? (especially for SIGSEGV, SIGBART, and SIGBUS)

qicosmos · 2021-06-01T06:09:22Z

@qicosmos can you post how the stacktrace looks like? Also does it make sense to write a python test to see if it prints logs on some signals? (especially for SIGSEGV, SIGBART, and SIGBUS)

signal_test will print the stacktrace, like this:

[ RUN ] SignalTest.SendBusSignalTest
*** SIGBUS received at time=1622527644 ***
[2021-06-01 14:07:24,263 E 99852 99852] signal_test.cc:42: SendBusSignalTest: kill pid 99866 with return value=0
[2021-06-01 14:07:24,263 E 99852 99852] signal_test.cc:42: SendBusSignalTest: kill pid 99866 with return value=0
PC: @ 0x5608e45461b2 (unknown) ray::TestSendSignal()
@ 0x7f624ca6e85d 64 absl::lts_2019_08_08::WriteFailureInfo()
@ 0x7f624ca6e9ff 96 absl::lts_2019_08_08::AbslFailureSignalHandler()
@ 0x7f623b73a5d0 246824784 (unknown)
@ 0x5608e45463fe 112 ray::SignalTest_SendBusSignalTest_Test::TestBody()
@ 0x7f623ba16475 48 testing::internal::HandleSehExceptionsInMethodIfSupported<>()
@ 0x7f623ba12127 144 testing::internal::HandleExceptionsInMethodIfSupported<>()
@ 0x7f623b9fe678 112 testing::Test::Run()
@ 0x7f623b9fefb0 112 testing::TestInfo::Run()
@ 0x7f623b9ff5fd 112 testing::TestCase::Run()
@ 0x7f623ba09783 112 testing::internal::UnitTestImpl::RunAllTests()
@ 0x7f623ba172fb 48 testing::internal::HandleSehExceptionsInMethodIfSupported<>()
@ 0x7f623ba12e85 144 testing::internal::HandleExceptionsInMethodIfSupported<>()
@ 0x7f623ba082ba 128 testing::UnitTest::Run()
@ 0x5608e45472b8 16 RUN_ALL_TESTS()
@ 0x5608e4546bbd 64 main
@ 0x7f623af64401 (unknown) __libc_start_main
@ 0x412e258d4c544155 (unknown) (unknown)

kfstorm · 2021-06-01T08:40:43Z

@qicosmos Could you also provide the output if we use glog? One thing I'm concerned about is that the thread ID is missing in your pasted output, so we cannot identify which thread receives/sends the signal. Thread ID is quite useful when we encounter SIGSEGV.

BTW, there's no assertion logic in signal_test.cc. We have to verify the stack trace output manually. Also, the test doesn't write logs to log files, which is what we also care for.

@qicosmos Could you run a simple Ray application and then kill the Raylet and take the stack trace logs from the Raylet log files?

kfstorm · 2021-06-01T08:43:00Z

bazel/ray_deps_setup.bzl

-        "query": split_on_query[1].split("&") if len(split_on_query) > 1 else None,
-        "fragment": split_on_anchor[1] if len(split_on_anchor) > 1 else None,
-    }
+  "scheme" : split_on_scheme[0],


Please revert unrelated changes in this file

kfstorm · 2021-06-01T08:44:50Z

src/ray/util/logging_test.cc

@@ -116,10 +116,13 @@ std::string TestFunctionLevel2() {
 #ifndef _WIN32
 TEST(PrintLogTest, CallstackTraceTest) {
  auto ret0 = TestFunctionLevel0();
+  std::cout << ret0 << std::endl;


Revert this file.

rkooo567

A couple last comments!

rkooo567 · 2021-06-02T05:42:36Z

src/ray/util/logging.cc

+  std::string output;
+  for (auto &stack : local_stack) {
+    if (absl::Symbolize(stack, buf, buf_size)) {
+      output.append("    ").append(buf).append("\n");


Hey is there any good way to add line information using addr2line here? https://linux.die.net/man/1/addr2line

I am just curious if that's possible since it looks like we can access each stack trace

rkooo567 · 2021-06-02T05:43:50Z

src/ray/util/logging.cc

-  google::InitGoogleLogging(app_name_.c_str());
-  int level = GetMappedSeverity(static_cast<RayLogLevel>(severity_threshold_));
-#endif
+#ifdef RAY_USE_SPDLOG


I am okay with just deleting glog to make iteration faster btw. (and we can remove Cerrlog in the separate PR)

rkooo567 · 2021-06-02T05:45:35Z

src/ray/util/logging_test.cc

  EXPECT_TRUE(ret1.find("TestFunctionLevel1") != std::string::npos);
  auto ret2 = TestFunctionLevel2();
+  std::cout << ret2 << std::endl;
  EXPECT_TRUE(ret2.find("TestFunctionLevel2") != std::string::npos);
 }


Can you add a python test? Maybe doing like

Start a raylet -> send a signal -> check log file -> make sure stacktrace is in the .err file

qicosmos · 2021-06-21T08:18:02Z

@qicosmos Could you also provide the output if we use glog? One thing I'm concerned about is that the thread ID is missing in your pasted output, so we cannot identify which thread receives/sends the signal. Thread ID is quite useful when we encounter SIGSEGV.

BTW, there's no assertion logic in signal_test.cc. We have to verify the stack trace output manually. Also, the test doesn't write logs to log files, which is what we also care for.

@qicosmos Could you run a simple Ray application and then kill the Raylet and take the stack trace logs from the Raylet log files?

This PR logging out:

[2021-06-21 16:13:38,663 E 15234 15234] signal_test.cc:42: SendBusSignalTest: kill pid 15236 with return value=0
[2021-06-21 16:13:38,663 E 15234 15234] signal_test.cc:42: SendBusSignalTest: kill pid 15236 with return value=0
PC: @     0x55eedf9a88c0  (unknown)  ray::TestSendSignal()
    @     0x7fc50d3704ed         80  absl::lts_2019_08_08::AbslFailureSignalHandler()
    @     0x7fc50c1a15d0  (unknown)  (unknown)
    @     0x55eedf9a8ad9         80  ray::SignalTest_SendBusSignalTest_Test::TestBody()
    @     0x7fc50c4113b7         96  testing::internal::HandleExceptionsInMethodIfSupported<>()
    @     0x7fc50c4115f2         48  testing::Test::Run()
    @     0x7fc50c4118e8         80  testing::TestInfo::Run()
    @     0x7fc50c411b75         80  testing::TestCase::Run()
    @     0x7fc50c411e18        144  testing::internal::UnitTestImpl::RunAllTests()
    @     0x7fc50c412127         96  testing::internal::HandleExceptionsInMethodIfSupported<>()
    @     0x7fc50c412322        128  testing::UnitTest::Run()
    @     0x55eedf9a7fd2        144  main
    @     0x7fc50b9c9401  (unknown)  __libc_start_main
    @ 0x44fe258d4c544155  (unknown)  (unknown)
[2021-06-21 16:13:38,667 E 15236 15234] logging.cc:299: *** SIGBUS received at time=1624263218 ***
[2021-06-21 16:13:38,667 E 15236 15234] logging.cc:299: *** SIGBUS received at time=1624263218 ***
[2021-06-21 16:13:38,667 E 15236 15234] logging.cc:299: PC: @     0x55eedf9a88c0  (unknown)  ray::TestSendSignal()
[2021-06-21 16:13:38,667 E 15236 15234] logging.cc:299: PC: @     0x55eedf9a88c0  (unknown)  ray::TestSendSignal()
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x7fc50d370516         80  absl::lts_2019_08_08::AbslFailureSignalHandler()
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x7fc50d370516         80  absl::lts_2019_08_08::AbslFailureSignalHandler()
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x7fc50c1a15d0  (unknown)  (unknown)
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x7fc50c1a15d0  (unknown)  (unknown)
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x55eedf9a8ad9         80  ray::SignalTest_SendBusSignalTest_Test::TestBody()
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x55eedf9a8ad9         80  ray::SignalTest_SendBusSignalTest_Test::TestBody()
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x7fc50c4113b7         96  testing::internal::HandleExceptionsInMethodIfSupported<>()
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x7fc50c4113b7         96  testing::internal::HandleExceptionsInMethodIfSupported<>()
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x7fc50c4115f2         48  testing::Test::Run()
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x7fc50c4115f2         48  testing::Test::Run()
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x7fc50c4118e8         80  testing::TestInfo::Run()
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x7fc50c4118e8         80  testing::TestInfo::Run()
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x7fc50c411b75         80  testing::TestCase::Run()
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x7fc50c411b75         80  testing::TestCase::Run()
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x7fc50c411e18        144  testing::internal::UnitTestImpl::RunAllTests()
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x7fc50c411e18        144  testing::internal::UnitTestImpl::RunAllTests()
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x7fc50c412127         96  testing::internal::HandleExceptionsInMethodIfSupported<>()
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x7fc50c412127         96  testing::internal::HandleExceptionsInMethodIfSupported<>()
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x7fc50c412322        128  testing::UnitTest::Run()
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x7fc50c412322        128  testing::UnitTest::Run()
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x55eedf9a7fd2        144  main
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x55eedf9a7fd2        144  main
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x7fc50b9c9401  (unknown)  __libc_start_main
[2021-06-21 16:13:38,668 E 15236 15234] logging.cc:299:     @     0x7fc50b9c9401  (unknown)  __libc_start_main
[2021-06-21 16:13:38,670 E 15236 15234] logging.cc:299:     @ 0x44fe258d4c544155  (unknown)  (unknown)
[2021-06-21 16:13:38,670 E 15236 15234] logging.cc:299:     @ 0x44fe258d4c544155  (unknown)  (unknown)
[       OK ] SignalTest.SendBusSignalTest (201 ms)
[ RUN      ] SignalTest.SIGABRT_Test

The glog logging out:

[2021-06-10 11:01:45,423 E 94936 94936] signal_test.cc:42: SendBusSignalTest: kill pid 94953 with return value=0
[2021-06-10 11:01:45,423 E 94936 94936] signal_test.cc:42: SendBusSignalTest: kill pid 94953 with return value=0
[2021-06-10 11:01:45,423 E 94953 94936] logging.cc:440: *** Aborted at 1623294105 (unix time) try "date -d @1623294105" if you are using GNU date ***
[2021-06-10 11:01:45,423 E 94953 94936] logging.cc:440: *** Aborted at 1623294105 (unix time) try "date -d @1623294105" if you are using GNU date ***
[2021-06-10 11:01:45,426 E 94953 94936] logging.cc:440: PC: @                0x0 (unknown)
[2021-06-10 11:01:45,426 E 94953 94936] logging.cc:440: PC: @                0x0 (unknown)
[2021-06-10 11:01:45,426 E 94953 94936] logging.cc:440: *** SIGBUS (@0x1f4000172d8) received by PID 94953 (TID 0x7f685213f940) from PID 94936; stack trace: ***
[2021-06-10 11:01:45,426 E 94953 94936] logging.cc:440: *** SIGBUS (@0x1f4000172d8) received by PID 94953 (TID 0x7f685213f940) from PID 94936; stack trace: ***
[2021-06-10 11:01:45,427 E 94953 94936] logging.cc:440:     @     0x7f6838efc5d0 (unknown)
[2021-06-10 11:01:45,427 E 94953 94936] logging.cc:440:     @     0x7f6838efc5d0 (unknown)
[2021-06-10 11:01:45,427 E 94953 94936] logging.cc:440:     @     0x561b23e24710 ray::TestSendSignal()
[2021-06-10 11:01:45,427 E 94953 94936] logging.cc:440:     @     0x561b23e24710 ray::TestSendSignal()
[2021-06-10 11:01:45,427 E 94953 94936] logging.cc:440:     @     0x561b23e24929 ray::SignalTest_SendBusSignalTest_Test::TestBody()
[2021-06-10 11:01:45,427 E 94953 94936] logging.cc:440:     @     0x561b23e24929 ray::SignalTest_SendBusSignalTest_Test::TestBody()
[2021-06-10 11:01:45,427 E 94953 94936] logging.cc:440:     @     0x7f683916ab67 testing::internal::HandleExceptionsInMethodIfSupported<>()
[2021-06-10 11:01:45,427 E 94953 94936] logging.cc:440:     @     0x7f683916ab67 testing::internal::HandleExceptionsInMethodIfSupported<>()
[2021-06-10 11:01:45,427 E 94953 94936] logging.cc:440:     @     0x7f683916ada2 testing::Test::Run()
[2021-06-10 11:01:45,427 E 94953 94936] logging.cc:440:     @     0x7f683916ada2 testing::Test::Run()
[2021-06-10 11:01:45,427 E 94953 94936] logging.cc:440:     @     0x7f683916b098 testing::TestInfo::Run()
[2021-06-10 11:01:45,427 E 94953 94936] logging.cc:440:     @     0x7f683916b098 testing::TestInfo::Run()
[2021-06-10 11:01:45,427 E 94953 94936] logging.cc:440:     @     0x7f683916b325 testing::TestCase::Run()
[2021-06-10 11:01:45,427 E 94953 94936] logging.cc:440:     @     0x7f683916b325 testing::TestCase::Run()
[2021-06-10 11:01:45,427 E 94953 94936] logging.cc:440:     @     0x7f683916b5c8 testing::internal::UnitTestImpl::RunAllTests()
[2021-06-10 11:01:45,427 E 94953 94936] logging.cc:440:     @     0x7f683916b5c8 testing::internal::UnitTestImpl::RunAllTests()
[2021-06-10 11:01:45,427 E 94953 94936] logging.cc:440:     @     0x7f683916b8d7 testing::internal::HandleExceptionsInMethodIfSupported<>()
[2021-06-10 11:01:45,427 E 94953 94936] logging.cc:440:     @     0x7f683916b8d7 testing::internal::HandleExceptionsInMethodIfSupported<>()
[2021-06-10 11:01:45,427 E 94953 94936] logging.cc:440:     @     0x7f683916bad2 testing::UnitTest::Run()
[2021-06-10 11:01:45,427 E 94953 94936] logging.cc:440:     @     0x7f683916bad2 testing::UnitTest::Run()
[2021-06-10 11:01:45,427 E 94953 94936] logging.cc:440:     @     0x561b23e23e52 main
[2021-06-10 11:01:45,427 E 94953 94936] logging.cc:440:     @     0x561b23e23e52 main
[2021-06-10 11:01:45,428 E 94953 94936] logging.cc:440:     @     0x7f6838726401 __libc_start_main
[2021-06-10 11:01:45,428 E 94953 94936] logging.cc:440:     @     0x7f6838726401 __libc_start_main
[2021-06-10 11:01:45,428 E 94953 94936] logging.cc:440:     @     0x561b23e23efa _start
[2021-06-10 11:01:45,428 E 94953 94936] logging.cc:440:     @     0x561b23e23efa _start
[       OK ] SignalTest.SendBusSignalTest (201 ms)
[ RUN      ] SignalTest.SIGABRT_Test

kfstorm · 2021-06-29T11:33:49Z

python/ray/tests/test_kill_raylet_signal_log.py

+    os.kill(pid, SIGKILL)
+    wait_for_pid_to_exit(pid)
+    thread = check_thread(raylet_out_path)
+    thread.start()


Forgot to join the thread.
Q1: Why do we need a thread here?
Q2: Will the exception thrown in the thread be populated to the main thread?

FYI: #16077 (comment)

You still didn't answer Q1 and Q2.

kfstorm · 2021-06-29T11:34:03Z

python/ray/tests/test_kill_raylet_signal_log.py

+class check_thread(threading.Thread):
+    def __init__(self, raylet_out_path):
+        threading.Thread.__init__(self)
+        self._daemonic = True


Why set this?

The _daemonic make the thread detach.
If we use thread.join() here, the log of SIGTERM can not be flushed, and we do not have a log flush interface, use detach thread can avoid the problem.

But in this case, the test will pass without waiting for the thread to finish. How do you ensure the signal is in the log file?

kfstorm · 2021-06-29T11:35:55Z

python/ray/tests/test_kill_raylet_signal_log.py

+        s = f.read()
+        assert len(s) > 0
+        print(s)
+        assert "SIGTERM" in s


For non-Windows environments, you send a SIGKILL to Raylet. Why do we always assert SIGTERM here?

Windows no SIGABRT, but has SIGTERM, use SIGTERM to support windows and linux.

For Linux and macOS, you send SIGKILL according to signal.SIGKILL if sys.platform != "win32" else signal.SIGTERM, right? But you assert SIGTERM here. So I suppose the test will fail on Linux and macOS.

kfstorm · 2021-07-08T11:21:15Z

@rkooo567 Any more comments?

qicosmos added 5 commits March 12, 2021 17:08

linkopts shared

c54e35f

Merge branch 'master' of https://github.com/ray-project/ray

8242f39

Merge branch 'master' of https://github.com/ray-project/ray

da20c29

Merge branch 'master' of https://github.com/ray-project/ray

be65dc1

remove glog

76eccb4

qicosmos requested review from kfstorm and ashione May 26, 2021 07:52

rkooo567 self-assigned this May 26, 2021

kfstorm reviewed May 26, 2021

View reviewed changes

Merge branch 'master' of https://github.com/ray-project/ray into remo…

9689252

…ve_glog_dependency

qicosmos added 2 commits May 31, 2021 14:17

install failure handler

cad0fda

Merge branch 'master' into remove_glog_dependency

c0e5fde

qicosmos requested a review from rkooo567 May 31, 2021 06:53

qicosmos added 3 commits May 31, 2021 15:36

show testing output

79311d8

set skipcount zero

5c4e3a3

fix

a15d2ed

kfstorm reviewed Jun 1, 2021

View reviewed changes

rkooo567 reviewed Jun 2, 2021

View reviewed changes

rkooo567 added the @author-action-required The PR author is responsible for the next step. Remove tag to send back to the reviewer. label Jun 2, 2021

qicosmos added 3 commits June 7, 2021 15:48

Merge branch 'master' of https://github.com/ray-project/ray

d2d0858

merge master

bdc8fab

revert some code; remove glog patches

0a266e2

qicosmos added 9 commits June 25, 2021 11:12

Merge branch 'master' into remove_glog_dependency

637bfd7

move to medium

b6ba295

support windows

c69df1f

Merge branch 'master' of https://github.com/ray-project/ray

2266433

fix for windows

42f49f6

Merge branch 'master' into remove_glog_dependency

3b030ca

lint

6ed519f

lint

d0fabea

Merge branch 'master' of https://github.com/ray-project/ray

1230cea

kfstorm reviewed Jun 29, 2021

View reviewed changes

qicosmos and others added 14 commits June 30, 2021 16:37

fix ut

b653678

Merge branch 'master' of https://github.com/ray-project/ray

37b5104

add wait timeout

95abfad

Merge branch 'master' into remove_glog_dependency

11a8840

Merge branch 'master' of https://github.com/ray-project/ray

e17ec36

Merge branch 'master' of https://github.com/ray-project/ray

fd62909

Merge branch 'master' into remove_glog_dependency

4c57c56

lint

f8431f3

Merge branch 'master' of https://github.com/ray-project/ray

2ccb6f5

Merge branch 'master' of https://github.com/ray-project/ray

f527d62

Merge branch 'master' of https://github.com/ray-project/ray

88630ed

Merge branch 'master' of https://github.com/ray-project/ray

72753f1

Merge branch 'master' of https://github.com/ray-project/ray

761d0bd

Merge branch 'master' into remove_glog_dependency

91bd83a

kfstorm approved these changes Jul 8, 2021

View reviewed changes

kfstorm removed the @author-action-required The PR author is responsible for the next step. Remove tag to send back to the reviewer. label Jul 8, 2021

rkooo567 approved these changes Jul 9, 2021

View reviewed changes

raulchen merged commit 298d2af into ray-project:master Jul 12, 2021

raulchen deleted the remove_glog_dependency branch July 12, 2021 09:06

	static constexpr size_t buf_size = 1 << 14;
	static constexpr size_t buf_size = 16 * 1024;

	/// This function return the current call stack information.
	/// This function returns the current call stack information.

[Ray Log] remove glog dependency #16077

[Ray Log] remove glog dependency #16077

Conversation

qicosmos commented May 26, 2021

Why are these changes needed?

Related issue number

Checks

kfstorm commented May 26, 2021

kfstorm commented May 26, 2021

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

kfstorm Jun 1, 2021 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

qicosmos commented May 26, 2021

rkooo567 commented May 29, 2021

rkooo567 commented Jun 1, 2021

qicosmos commented Jun 1, 2021 • edited Loading

kfstorm commented Jun 1, 2021

Choose a reason for hiding this comment

Choose a reason for hiding this comment

rkooo567 left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

qicosmos commented Jun 21, 2021

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

qicosmos Jun 29, 2021 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

kfstorm commented Jul 8, 2021

kfstorm Jun 1, 2021 •

edited

Loading

qicosmos commented Jun 1, 2021 •

edited

Loading

qicosmos Jun 29, 2021 •

edited

Loading