PermanentlyFailedException: Timed out waiting for service to come alive.

[ghost]

I am trying out the demo code for using the CoreNLP server.

# set up the client
with CoreNLPClient(annotators=['tokenize','ssplit','pos','lemma','ner','parse','depparse','coref'], timeout=30000, memory='16G') as client:
    # submit the request to the server
    ann = client.annotate(text)

gives me this error:

PermanentlyFailedException: Timed out waiting for service to come alive.

I tried increasing the timeout limit, but no success.

[yuhaozhang]

Did you download and set up the CoreNLP package by following this setup guide?

[fpcsong]

I have the same problem each time after annotated 2000+ sentences, you can find if there is some thing overflow upd: It's true, after I limit run times of each annotator instance's, it can work well.

if client is None or annotate_time > 1000:
        client = None
        time.sleep(5)
        client = CoreNLPClient(annotators=['ssplit', 'tokenize'], memory='8G', timeout=30000)
        annotate_time=0

[ghost]

Did you download and set up the CoreNLP package by following this setup guide?

Yes. I followed the setup guide and specified the path in my code.

[yuhaozhang]

I just tried the running the same code on a Ubuntu system with 10000 test sentences and everything works fine for me. It is hard to find what's going on without further information.

@sudeshna-d Can you try initializing the CoreNLPClient with be_quiet=False argument and rerun your example? This will allow the underlying CoreNLP logs to be printed to stderr. We may be able to find out what's going with the CoreNLP logs. Something like this should work:

# set up the client
with CoreNLPClient(annotators=['tokenize','ssplit','pos','lemma','ner','parse','depparse','coref'], timeout=30000, memory='16G', be_quiet=False) as client:
    # other code

@fpcsong It is also likely that a specific sentence crashed the CoreNLP server, which then generated this PermanentlyFailedException for you. Can you try:

1. Add the be_quiet=False argument as detailed above and rerun your code;
2. Try to print out the example which the program crashed on?

[ghost]

@yuhaozhang

This is the error that I get on using the be_quiet=False argument:

---------------------------------------------------------------------------
UnsupportedOperation                      Traceback (most recent call last)
<ipython-input-5-7569e3faaa93> in <module>
     18 
     19 # set up the client
---> 20 with CoreNLPClient(annotators=['tokenize','ssplit','pos','lemma','ner','parse','depparse','coref'], timeout=30000, memory='16G',be_quiet=False) as client:
     21     # submit the request to the server
     22     ann = client.annotate(text)

~/envs/.env/lib/python3.6/site-packages/stanfordnlp/server/client.py in __enter__(self)
     87 
     88     def __enter__(self):
---> 89         self.start()
     90         return self
     91 

~/envs/.env/lib/python3.6/site-packages/stanfordnlp/server/client.py in start(self)
     77             self.server = subprocess.Popen(self.start_cmd,
     78                                            stderr=stderr,
---> 79                                            stdout=stderr)
     80 
     81     def stop(self):

/usr/lib/python3.6/subprocess.py in __init__(self, args, bufsize, executable, stdin, stdout, stderr, preexec_fn, close_fds, shell, cwd, env, universal_newlines, startupinfo, creationflags, restore_signals, start_new_session, pass_fds, encoding, errors)
    685         (p2cread, p2cwrite,
    686          c2pread, c2pwrite,
--> 687          errread, errwrite) = self._get_handles(stdin, stdout, stderr)
    688 
    689         # We wrap OS handles *before* launching the child, otherwise a

/usr/lib/python3.6/subprocess.py in _get_handles(self, stdin, stdout, stderr)
   1202             else:
   1203                 # Assuming file-like object
-> 1204                 c2pwrite = stdout.fileno()
   1205 
   1206             if stderr is None:

UnsupportedOperation: fileno

[yuhaozhang]

@sudeshna-d That's weird. It seems like the issue is not even in stanfordnlp, but a call to the native python subprocess library. Seems like a fileno() function call is not supported for your stdout object.

Are you running python directly from your terminal or an IDE. If the latter, seems like someone has reported a similar issue here: https://stackoverflow.com/questions/31080829/python-error-io-unsupportedoperation-fileno.

[ghost]

@yuhaozhang I am using Jupyter Notebook. The demo code does work when used directly from terminal. Thanks for all the help!

[suganth1997]

@sudeshna-d If you want to run in Jupyter Notebook execute without the with block.

[Suki-m730026038]

Hi, I suffer the same problem . Have you already solved this problem?