def g():
        s = Scheduler(ip='127.0.0.1')
        done = s.start()
        s.listen(0)
        a = Worker('127.0.0.1', s.port, ncores=2, ip='127.0.0.1')
        yield a._start()
        b = Worker('127.0.0.1', s.port, ncores=1, ip=b_ip)
        yield b._start()

        start = time()
        try:
            while len(s.ncores) < 2:
                yield gen.sleep(0.01)
                if time() - start > 5:
                    raise Exception("Cluster creation timeout")

            yield f(s, a, b)
        finally:
            logger.debug("Closing out test cluster")
            for w in [a, b]:
                with ignoring(TimeoutError, StreamClosedError, OSError):
                    yield w._close()
                if os.path.exists(w.local_dir):
                    shutil.rmtree(w.local_dir)
            yield s.close()
Beispiel #2
0
def main(center, host, port, http_port, bokeh_port, show, _bokeh, bokeh_whitelist):
    given_host = host
    host = host or get_ip()
    ip = socket.gethostbyname(host)
    loop = IOLoop.current()
    scheduler = Scheduler(center, ip=ip,
                          services={('http', http_port): HTTPScheduler})
    if center:
        loop.run_sync(scheduler.sync_center)
    scheduler.start(port)

    if _bokeh:
        try:
            import bokeh
            import distributed.bokeh
            hosts = ['%s:%d' % (h, bokeh_port) for h in
                     ['localhost', '127.0.0.1', ip, socket.gethostname(),
                      host] + list(bokeh_whitelist)]
            dirname = os.path.dirname(distributed.__file__)
            paths = [os.path.join(dirname, 'bokeh', name)
                     for name in ['status', 'tasks']]
            binname = sys.argv[0][:-len('dscheduler')] + 'bokeh'
            args = ([binname, 'serve'] + paths +
                    ['--log-level', 'warning',
                     '--check-unused-sessions=50',
                     '--unused-session-lifetime=1',
                     '--port', str(bokeh_port)] +
                     sum([['--host', host] for host in hosts], []))
            if show:
                args.append('--show')

            bokeh_options = {'host': host if given_host else '127.0.0.1',
                             'http-port': http_port,
                             'tcp-port': port,
                             'bokeh-port': bokeh_port}
            with open('.dask-web-ui.json', 'w') as f:
                json.dump(bokeh_options, f, indent=2)

            if sys.version_info[0] >= 3:
                from bokeh.command.bootstrap import main
                ctx = multiprocessing.get_context('spawn')
                bokeh_proc = ctx.Process(target=main, args=(args,))
                bokeh_proc.daemon = True
                bokeh_proc.start()
            else:
                bokeh_proc = subprocess.Popen(args)

            logger.info(" Bokeh UI at:  http://%s:%d/status/"
                        % (ip, bokeh_port))
        except ImportError:
            logger.info("Please install Bokeh to get Web UI")
        except Exception as e:
            logger.warn("Could not start Bokeh web UI", exc_info=True)

    loop.start()
    loop.close()
    scheduler.stop()
    bokeh_proc.terminate()

    logger.info("End scheduler at %s:%d", ip, port)
Beispiel #3
0
def start_cluster(ncores, scheduler_addr, loop, security=None,
                  Worker=Worker, scheduler_kwargs={}, worker_kwargs={}):
    s = Scheduler(loop=loop, validate=True, security=security,
                  **scheduler_kwargs)
    done = s.start(scheduler_addr)
    workers = [Worker(s.address, ncores=ncore[1], name=i, security=security,
                      loop=loop, validate=True,
                      **(merge(worker_kwargs, ncore[2])
                         if len(ncore) > 2
                         else worker_kwargs))
               for i, ncore in enumerate(ncores)]
    for w in workers:
        w.rpc = workers[0].rpc

    yield [w._start(ncore[0]) for ncore, w in zip(ncores, workers)]

    start = time()
    while (len(s.workers) < len(ncores) or
           any(comm.comm is None for comm in s.stream_comms.values())):
        yield gen.sleep(0.01)
        if time() - start > 5:
            yield [w._close(timeout=1) for w in workers]
            yield s.close(fast=True)
            raise Exception("Cluster creation timeout")
    raise gen.Return((s, workers))
Beispiel #4
0
def main(center, host, port, http_port, bokeh_port, show, _bokeh, bokeh_whitelist):
    given_host = host
    host = host or get_ip()
    if ':' in host and port == 8786:
        host, port = host.rsplit(':', 1)
        port = int(port)
    ip = socket.gethostbyname(host)
    loop = IOLoop.current()
    scheduler = Scheduler(center, ip=ip,
                          services={('http', http_port): HTTPScheduler},
                          loop=loop)
    scheduler.start(port)

    bokeh_proc = None
    if _bokeh:
        try:
            from distributed.bokeh.application import BokehWebInterface
            bokeh_proc = BokehWebInterface(host=host, http_port=http_port,
                    tcp_port=port, bokeh_port=bokeh_port,
                    bokeh_whitelist=bokeh_whitelist, show=show)
        except ImportError:
            logger.info("Please install Bokeh to get Web UI")
        except Exception as e:
            logger.warn("Could not start Bokeh web UI", exc_info=True)

    loop.start()
    loop.close()
    scheduler.stop()
    if bokeh_proc:
        bokeh_proc.close()

    logger.info("End scheduler at %s:%d", ip, port)
Beispiel #5
0
def main(center, host, port, http_port, bokeh_port):
    ip = socket.gethostbyname(host)
    loop = IOLoop.current()
    scheduler = Scheduler(center, ip=ip,
                          services={('http', http_port): HTTPScheduler})
    if center:
        loop.run_sync(scheduler.sync_center)
    scheduler.start(port)

    try:
        import bokeh
    except ImportError:
        pass
    else:
        import distributed.diagnostics.bokeh
        hosts = ['%s:%d' % (h, bokeh_port) for h in
                 ['localhost', '127.0.0.1', ip, socket.gethostname()]]
        dirname = os.path.dirname(distributed.__file__)
        path = os.path.join(dirname, 'diagnostics', 'bokeh', 'status')
        proc = subprocess.Popen(['bokeh', 'serve', path,
                                 '--log-level', 'warning',
                                 '--port', str(bokeh_port)] +
                                 sum([['--host', host] for host in hosts], []))

        distributed.diagnostics.bokeh.server_process = proc  # monkey patch

        logger.info(" Start Bokeh UI at:        http://%s:%d/status/"
                    % (ip, bokeh_port))

    loop.start()
    loop.close()
    scheduler.stop()

    logger.info("End scheduler at %s:%d", ip, port)
Beispiel #6
0
def test_scheduler_file():
    with tmpfile() as fn:
        s = Scheduler(scheduler_file=fn)
        s.start(8008)
        w = Nanny(scheduler_file=fn)
        yield w._start()
        assert set(s.workers) == {w.worker_address}
        yield w._close()
        s.stop()
Beispiel #7
0
def main(scheduler_file, interface, nthreads, local_directory, memory_limit,
         scheduler, bokeh_port, bokeh_prefix, nanny, bokeh_worker_port):
    if interface:
        host = get_ip_interface(interface)
    else:
        host = None

    if rank == 0 and scheduler:
        try:
            from distributed.bokeh.scheduler import BokehScheduler
        except ImportError:
            services = {}
        else:
            services = {('bokeh',  bokeh_port): partial(BokehScheduler,
                                                        prefix=bokeh_prefix)}
        scheduler = Scheduler(scheduler_file=scheduler_file,
                              loop=loop,
                              services=services)
        addr = uri_from_host_port(host, None, 8786)
        scheduler.start(addr)
        try:
            loop.start()
            loop.close()
        finally:
            scheduler.stop()
    else:
        W = Nanny if nanny else Worker
        worker = W(scheduler_file=scheduler_file,
                   loop=loop,
                   name=rank if scheduler else None,
                   ncores=nthreads,
                   local_dir=local_directory,
                   services={('bokeh', bokeh_worker_port): BokehWorker},
                   memory_limit=memory_limit)
        addr = uri_from_host_port(host, None, 0)

        @gen.coroutine
        def run():
            yield worker._start(addr)
            while worker.status != 'closed':
                yield gen.sleep(0.2)

        try:
            loop.run_sync(run)
            loop.close()
        finally:
            pass

        @gen.coroutine
        def close():
            yield worker._close(timeout=2)

        loop.run_sync(close)
Beispiel #8
0
def main(center, port):
    loop = IOLoop.current()
    scheduler = Scheduler(center, services={'http': HTTPScheduler})
    if center:
        loop.run_sync(scheduler.sync_center)
    done = scheduler.start(port)

    loop.start()
    loop.close()
    scheduler.stop()

    logging.info("End scheduler at %s:%d", ip, port)
Beispiel #9
0
def start_cluster(ncores, Worker=Worker):
    s = Scheduler(ip='127.0.0.1')
    done = s.start(0)
    workers = [Worker(s.ip, s.port, ncores=v, ip=k) for k, v in ncores]

    yield [w._start() for w in workers]

    start = time()
    while len(s.ncores) < len(ncores):
        yield gen.sleep(0.01)
        if time() - start > 5:
            raise Exception("Cluster creation timeout")
    raise gen.Return((s, workers))
def main(center, host, port):
    ip = socket.gethostbyname(host)
    loop = IOLoop.current()
    scheduler = Scheduler(center, services={'http': HTTPScheduler}, ip=ip)
    if center:
        loop.run_sync(scheduler.sync_center)
    scheduler.start(port)

    loop.start()
    loop.close()
    scheduler.stop()

    logging.info("End scheduler at %s:%d", ip, port)
Beispiel #11
0
def start_cluster(ncores):
    s = Scheduler(ip='127.0.0.1')
    s.listen(0)
    workers = [Worker(s.ip, s.port, ncores=v, ip=k) for k, v in ncores]

    IOLoop.current().add_callback(s.start)
    yield [w._start() for w in workers]

    start = time()
    while len(s.ncores) < len(ncores):
        yield gen.sleep(0.01)
        if time() - start > 5:
            raise Exception("Cluster creation timeout")
    raise gen.Return((s, workers))
Beispiel #12
0
def start_cluster(ncores, loop, Worker=Worker):
    s = Scheduler(ip='127.0.0.1', loop=loop)
    done = s.start(0)
    workers = [Worker(s.ip, s.port, ncores=v, ip=k, name=i, loop=loop)
                for i, (k, v) in enumerate(ncores)]

    yield [w._start() for w in workers]

    start = time()
    while len(s.ncores) < len(ncores):
        yield gen.sleep(0.01)
        if time() - start > 5:
            raise Exception("Cluster creation timeout")
    raise gen.Return((s, workers))
    def f(c, a, b):
        s = Scheduler((c.ip, c.port), loop=loop)
        yield s.sync_center()
        done = s.start(0)

        progress = TextProgressBar([], scheduler=(s.ip, s.port), start=False,
                                   interval=0.01)
        yield progress.listen()

        assert progress.status == 'finished'
        check_bar_completed(capsys)

        s.close()
        yield done
Beispiel #14
0
def run_scheduler(q, nputs, **kwargs):
    from distributed import Scheduler

    # On Python 2.7 and Unix, fork() is used to spawn child processes,
    # so avoid inheriting the parent's IO loop.
    with pristine_loop() as loop:
        scheduler = Scheduler(validate=True, **kwargs)
        done = scheduler.start('127.0.0.1')

        for i in range(nputs):
            q.put(scheduler.address)
        try:
            loop.start()
        finally:
            loop.close(all_fds=True)
Beispiel #15
0
def run_scheduler(q, scheduler_port=0, **kwargs):
    from distributed import Scheduler
    from tornado.ioloop import IOLoop, PeriodicCallback
    IOLoop.clear_instance()
    loop = IOLoop(); loop.make_current()
    PeriodicCallback(lambda: None, 500).start()

    scheduler = Scheduler(loop=loop, validate=True, **kwargs)
    done = scheduler.start(scheduler_port)

    q.put(scheduler.port)
    try:
        loop.start()
    finally:
        loop.close(all_fds=True)
def main(host, port, http_port, bokeh_port, show, _bokeh,
         bokeh_whitelist, prefix, use_xheaders, pid_file):

    if pid_file:
        with open(pid_file, 'w') as f:
            f.write(str(os.getpid()))

        def del_pid_file():
            if os.path.exists(pid_file):
                os.remove(pid_file)
        atexit.register(del_pid_file)

    soft, hard = resource.getrlimit(resource.RLIMIT_NOFILE)
    limit = max(soft, hard // 2)
    resource.setrlimit(resource.RLIMIT_NOFILE, (limit, hard))

    given_host = host
    host = host or get_ip()
    if ':' in host and port == 8786:
        host, port = host.rsplit(':', 1)
        port = int(port)
    ip = socket.gethostbyname(host)
    loop = IOLoop.current()
    scheduler = Scheduler(ip=ip, loop=loop,
                          services={('http', http_port): HTTPScheduler})
    scheduler.start(port)

    bokeh_proc = None
    if _bokeh:
        try:
            from distributed.bokeh.application import BokehWebInterface
            bokeh_proc = BokehWebInterface(host=host, http_port=http_port,
                    tcp_port=port, bokeh_port=bokeh_port,
                    bokeh_whitelist=bokeh_whitelist, show=show, prefix=prefix,
                    use_xheaders=use_xheaders, quiet=False)
        except ImportError:
            logger.info("Please install Bokeh to get Web UI")
        except Exception as e:
            logger.warn("Could not start Bokeh web UI", exc_info=True)

    loop.start()
    loop.close()
    scheduler.stop()
    if bokeh_proc:
        bokeh_proc.close()

    logger.info("End scheduler at %s:%d", ip, port)
Beispiel #17
0
    def f():
        s = Scheduler(loop=loop)
        done = s.start(0)
        a = Worker(s.ip, s.port, loop=loop, ncores=1)
        b = Worker(s.ip, s.port, loop=loop, ncores=1)
        yield [a._start(0), b._start(0)]

        progress = TextProgressBar([], scheduler=(s.ip, s.port), start=False,
                                   interval=0.01)
        yield progress.listen()

        assert progress.status == 'finished'
        check_bar_completed(capsys)

        yield [a._close(), b._close()]
        s.close()
        yield done
Beispiel #18
0
def run_scheduler(q, scheduler_port=0, center_port=None, **kwargs):
    from distributed import Scheduler
    from tornado.ioloop import IOLoop, PeriodicCallback
    import logging
    IOLoop.clear_instance()
    loop = IOLoop(); loop.make_current()
    PeriodicCallback(lambda: None, 500).start()
    logging.getLogger("tornado").setLevel(logging.CRITICAL)

    center = ('127.0.0.1', center_port) if center_port else None
    scheduler = Scheduler(center=center, loop=loop, validate=True, **kwargs)
    done = scheduler.start(scheduler_port)

    q.put(scheduler.port)
    try:
        loop.start()
    finally:
        loop.close(all_fds=True)
Beispiel #19
0
def main(center, host, port, http_port, bokeh_port, show, _bokeh):
    ip = socket.gethostbyname(host)
    loop = IOLoop.current()
    scheduler = Scheduler(center, ip=ip,
                          services={('http', http_port): HTTPScheduler})
    if center:
        loop.run_sync(scheduler.sync_center)
    scheduler.start(port)

    if _bokeh:
        try:
            import bokeh
            import distributed.bokeh
            hosts = ['%s:%d' % (h, bokeh_port) for h in
                     ['localhost', '127.0.0.1', ip, socket.gethostname(), host]]
            dirname = os.path.dirname(distributed.__file__)
            paths = [os.path.join(dirname, 'bokeh', name)
                     for name in ['status', 'tasks']]
            args = (['bokeh', 'serve'] + paths +
                    ['--log-level', 'warning',
                     '--check-unused-sessions=50',
                     '--unused-session-lifetime=1',
                     '--port', str(bokeh_port)] +
                     sum([['--host', host] for host in hosts], []))
            if show:
                args.append('--show')
            from bokeh.command.bootstrap import main
            proc = multiprocessing.Process(target=main, args=(args,))
            proc.daemon = True
            proc.start()

            logger.info(" Start Bokeh UI at:        http://%s:%d/status/"
                        % (ip, bokeh_port))
        except ImportError:
            logger.info("Please install Bokeh to get Web UI")
        except Exception as e:
            logger.warn("Could not start Bokeh web UI", exc_info=True)

    loop.start()
    loop.close()
    scheduler.stop()
    proc.terminate()

    logger.info("End scheduler at %s:%d", ip, port)
Beispiel #20
0
def run_scheduler(q, center_port=None, **kwargs):
    from distributed import Scheduler
    from tornado.ioloop import IOLoop, PeriodicCallback
    import logging
    IOLoop.clear_instance()
    loop = IOLoop(); loop.make_current()
    PeriodicCallback(lambda: None, 500).start()
    logging.getLogger("tornado").setLevel(logging.CRITICAL)

    center = ('127.0.0.1', center_port) if center_port else None
    scheduler = Scheduler(center=center, **kwargs)
    scheduler.listen(0)

    if center_port:
        loop.run_sync(scheduler.sync_center)
    done = scheduler.start(0)

    q.put(scheduler.port)
    loop.start()
Beispiel #21
0
def start_cluster(ncores, loop, Worker=Worker, scheduler_kwargs={},
                  worker_kwargs={}):
    s = Scheduler(ip='127.0.0.1', loop=loop, validate=True, **scheduler_kwargs)
    done = s.start(0)
    workers = [Worker(s.ip, s.port, ncores=ncore[1], ip=ncore[0], name=i,
                      loop=loop, validate=True,
                      **(merge(worker_kwargs, ncore[2])
                         if len(ncore) > 2
                         else worker_kwargs))
                for i, ncore in enumerate(ncores)]
    for w in workers:
        w.rpc = workers[0].rpc

    yield [w._start() for w in workers]

    start = time()
    while len(s.ncores) < len(ncores):
        yield gen.sleep(0.01)
        if time() - start > 5:
            raise Exception("Cluster creation timeout")
    raise gen.Return((s, workers))
    def f(c, a, b):
        s = Scheduler((c.ip, c.port), loop=loop)
        s.listen(0)
        yield s.sync_center()
        done = s.start()

        s.update_graph(dsk={'x': (div, 1, 0)},
                       keys=['x'])

        progress = TextProgressBar(['x'], scheduler=(s.ip, s.port),
                                   start=False, interval=0.01)
        yield progress.listen()

        assert progress.status == 'error'
        assert progress.stream.closed()

        progress = TextProgressBar(['x'], scheduler=(s.ip, s.port),
                                   start=False, interval=0.01)
        yield progress.listen()
        assert progress.status == 'error'
        assert progress.stream.closed()

        s.close()
        yield done
Beispiel #23
0
def main(host, port, http_port, bokeh_port, show, _bokeh, bokeh_whitelist,
         prefix, use_xheaders, pid_file):

    if pid_file:
        with open(pid_file, 'w') as f:
            f.write(str(os.getpid()))

        def del_pid_file():
            if os.path.exists(pid_file):
                os.remove(pid_file)

        atexit.register(del_pid_file)

    if sys.platform.startswith('linux'):
        import resource  # module fails importing on Windows
        soft, hard = resource.getrlimit(resource.RLIMIT_NOFILE)
        limit = max(soft, hard // 2)
        resource.setrlimit(resource.RLIMIT_NOFILE, (limit, hard))

    given_host = host
    host = host or get_ip()
    if ':' in host and port == 8786:
        host, port = host.rsplit(':', 1)
        port = int(port)
    ip = socket.gethostbyname(host)
    loop = IOLoop.current()
    logger.info('-' * 47)

    services = {('http', http_port): HTTPScheduler}
    if _bokeh:
        from distributed.bokeh.scheduler import BokehScheduler
        services[('bokeh', 8788)] = BokehScheduler
    scheduler = Scheduler(ip=ip, loop=loop, services=services)
    scheduler.start(port)

    bokeh_proc = None
    if _bokeh:
        try:
            from distributed.bokeh.application import BokehWebInterface
            bokeh_proc = BokehWebInterface(host=host,
                                           http_port=http_port,
                                           tcp_port=port,
                                           bokeh_port=bokeh_port,
                                           bokeh_whitelist=bokeh_whitelist,
                                           show=show,
                                           prefix=prefix,
                                           use_xheaders=use_xheaders,
                                           quiet=False)
        except ImportError:
            logger.info("Please install Bokeh to get Web UI")
        except Exception as e:
            logger.warn("Could not start Bokeh web UI", exc_info=True)

    logger.info('-' * 47)
    try:
        loop.start()
        loop.close()
    finally:
        scheduler.stop()
        if bokeh_proc:
            bokeh_proc.close()

        logger.info("End scheduler at %s:%d", ip, port)
Beispiel #24
0
async def test_web_preload_worker(cleanup, worker_preload):
    async with Scheduler(port=8786, host="localhost") as s:
        async with Nanny(
                preload_nanny=["http://127.0.0.1:12346/preload"]) as nanny:
            assert nanny.scheduler_addr == s.address
Beispiel #25
0
def main(
    host,
    port,
    bokeh_port,
    show,
    _bokeh,
    bokeh_whitelist,
    bokeh_prefix,
    use_xheaders,
    pid_file,
    scheduler_file,
    interface,
    local_directory,
    preload,
    preload_argv,
    tls_ca_file,
    tls_cert,
    tls_key,
    dashboard_address,
):
    g0, g1, g2 = gc.get_threshold()  # https://github.com/dask/distributed/issues/1653
    gc.set_threshold(g0 * 3, g1 * 3, g2 * 3)

    enable_proctitle_on_current()
    enable_proctitle_on_children()

    if bokeh_port is not None:
        warnings.warn(
            "The --bokeh-port flag has been renamed to --dashboard-address. "
            "Consider adding ``--dashboard-address :%d`` " % bokeh_port
        )
        dashboard_address = bokeh_port

    sec = Security(
        tls_ca_file=tls_ca_file, tls_scheduler_cert=tls_cert, tls_scheduler_key=tls_key
    )

    if not host and (tls_ca_file or tls_cert or tls_key):
        host = "tls://"

    if pid_file:
        with open(pid_file, "w") as f:
            f.write(str(os.getpid()))

        def del_pid_file():
            if os.path.exists(pid_file):
                os.remove(pid_file)

        atexit.register(del_pid_file)

    local_directory_created = False
    if local_directory:
        if not os.path.exists(local_directory):
            os.mkdir(local_directory)
            local_directory_created = True
    else:
        local_directory = tempfile.mkdtemp(prefix="scheduler-")
        local_directory_created = True
    if local_directory not in sys.path:
        sys.path.insert(0, local_directory)

    if sys.platform.startswith("linux"):
        import resource  # module fails importing on Windows

        soft, hard = resource.getrlimit(resource.RLIMIT_NOFILE)
        limit = max(soft, hard // 2)
        resource.setrlimit(resource.RLIMIT_NOFILE, (limit, hard))

    if interface:
        if host:
            raise ValueError("Can not specify both interface and host")
        else:
            host = get_ip_interface(interface)

    addr = uri_from_host_port(host, port, 8786)

    loop = IOLoop.current()
    logger.info("-" * 47)

    services = {}
    if _bokeh:
        try:
            from distributed.bokeh.scheduler import BokehScheduler

            services[("bokeh", dashboard_address)] = (
                BokehScheduler,
                {"prefix": bokeh_prefix},
            )
        except ImportError as error:
            if str(error).startswith("No module named"):
                logger.info("Web dashboard not loaded.  Unable to import bokeh")
            else:
                logger.info("Unable to import bokeh: %s" % str(error))

    scheduler = Scheduler(
        loop=loop, services=services, scheduler_file=scheduler_file, security=sec
    )
    scheduler.start(addr)
    if not preload:
        preload = dask.config.get("distributed.scheduler.preload")
    if not preload_argv:
        preload_argv = dask.config.get("distributed.scheduler.preload-argv")
    preload_modules(
        preload, parameter=scheduler, file_dir=local_directory, argv=preload_argv
    )

    logger.info("Local Directory: %26s", local_directory)
    logger.info("-" * 47)

    install_signal_handlers(loop)

    try:
        loop.start()
        loop.close()
    finally:
        scheduler.stop()
        if local_directory_created:
            shutil.rmtree(local_directory)

        logger.info("End scheduler at %r", addr)
Beispiel #26
0
def main(host, port, bokeh_port, show, _bokeh, bokeh_whitelist, bokeh_prefix,
        use_xheaders, pid_file, scheduler_file, interface,
        local_directory, preload, preload_argv, tls_ca_file, tls_cert, tls_key):

    enable_proctitle_on_current()
    enable_proctitle_on_children()

    sec = Security(tls_ca_file=tls_ca_file,
                   tls_scheduler_cert=tls_cert,
                   tls_scheduler_key=tls_key,
                   )

    if not host and (tls_ca_file or tls_cert or tls_key):
        host = 'tls://'

    if pid_file:
        with open(pid_file, 'w') as f:
            f.write(str(os.getpid()))

        def del_pid_file():
            if os.path.exists(pid_file):
                os.remove(pid_file)
        atexit.register(del_pid_file)

    local_directory_created = False
    if local_directory:
        if not os.path.exists(local_directory):
            os.mkdir(local_directory)
            local_directory_created = True
    else:
        local_directory = tempfile.mkdtemp(prefix='scheduler-')
        local_directory_created = True
    if local_directory not in sys.path:
        sys.path.insert(0, local_directory)

    if sys.platform.startswith('linux'):
        import resource   # module fails importing on Windows
        soft, hard = resource.getrlimit(resource.RLIMIT_NOFILE)
        limit = max(soft, hard // 2)
        resource.setrlimit(resource.RLIMIT_NOFILE, (limit, hard))

    if interface:
        if host:
            raise ValueError("Can not specify both interface and host")
        else:
            host = get_ip_interface(interface)

    addr = uri_from_host_port(host, port, 8786)

    loop = IOLoop.current()
    logger.info('-' * 47)

    services = {}
    if _bokeh:
        try:
            from distributed.bokeh.scheduler import BokehScheduler
            services[('bokeh', bokeh_port)] = (BokehScheduler,
                                               {'prefix': bokeh_prefix})
        except ImportError as error:
            if str(error).startswith('No module named'):
                logger.info('Web dashboard not loaded.  Unable to import bokeh')
            else:
                logger.info('Unable to import bokeh: %s' % str(error))

    scheduler = Scheduler(loop=loop, services=services,
                          scheduler_file=scheduler_file,
                          security=sec)
    scheduler.start(addr)
    if not preload:
        preload = dask.config.get('distributed.scheduler.preload')
    if not preload_argv:
        preload_argv = dask.config.get('distributed.scheduler.preload-argv')
    preload_modules(preload, parameter=scheduler, file_dir=local_directory, argv=preload_argv)

    logger.info('Local Directory: %26s', local_directory)
    logger.info('-' * 47)

    install_signal_handlers(loop)

    try:
        loop.start()
        loop.close()
    finally:
        scheduler.stop()
        if local_directory_created:
            shutil.rmtree(local_directory)

        logger.info("End scheduler at %r", addr)
def test_services():
    s = Scheduler(services={'http': HTTPScheduler})
    assert isinstance(s.services['http'], HTTPServer)
    assert s.services['http'].port
def test_services_with_port():
    s = Scheduler(services={('http', 9999): HTTPScheduler})
    assert isinstance(s.services['http'], HTTPServer)
    assert s.services['http'].port == 9999
async def test_tls_scheduler(security, cleanup):
    async with Scheduler(security=security, host="localhost") as s:
        assert s.address.startswith("tls")
Beispiel #30
0
async def test_config(cleanup):
    async with Scheduler() as s:
        async with Nanny(s.address, config={"foo": "bar"}) as n:
            async with Client(s.address, asynchronous=True) as client:
                config = await client.run(dask.config.get, "foo")
                assert config[n.worker_address] == "bar"
Beispiel #31
0
 def _create_scheduler(self):
     scheduler = Scheduler(ip=self.ip)
     scheduler.start(self.port)
     return scheduler
Beispiel #32
0
def main(host, port, http_port, bokeh_port, show, _bokeh, bokeh_whitelist, prefix, use_xheaders, pid_file):

    if pid_file:
        with open(pid_file, "w") as f:
            f.write(str(os.getpid()))

        def del_pid_file():
            if os.path.exists(pid_file):
                os.remove(pid_file)

        atexit.register(del_pid_file)

    if sys.platform.startswith("linux"):
        import resource  # module fails importing on Windows

        soft, hard = resource.getrlimit(resource.RLIMIT_NOFILE)
        limit = max(soft, hard // 2)
        resource.setrlimit(resource.RLIMIT_NOFILE, (limit, hard))

    given_host = host
    host = host or get_ip()
    if ":" in host and port == 8786:
        host, port = host.rsplit(":", 1)
        port = int(port)
    ip = socket.gethostbyname(host)
    loop = IOLoop.current()
    logger.info("-" * 47)

    services = {("http", http_port): HTTPScheduler}
    if _bokeh:
        from distributed.bokeh.scheduler import BokehScheduler

        services[("bokeh", 8788)] = BokehScheduler
    scheduler = Scheduler(ip=ip, loop=loop, services=services)
    scheduler.start(port)

    bokeh_proc = None
    if _bokeh:
        try:
            from distributed.bokeh.application import BokehWebInterface

            bokeh_proc = BokehWebInterface(
                host=host,
                http_port=http_port,
                tcp_port=port,
                bokeh_port=bokeh_port,
                bokeh_whitelist=bokeh_whitelist,
                show=show,
                prefix=prefix,
                use_xheaders=use_xheaders,
                quiet=False,
            )
        except ImportError:
            logger.info("Please install Bokeh to get Web UI")
        except Exception as e:
            logger.warn("Could not start Bokeh web UI", exc_info=True)

    logger.info("-" * 47)
    try:
        loop.start()
        loop.close()
    finally:
        scheduler.stop()
        if bokeh_proc:
            bokeh_proc.close()

        logger.info("End scheduler at %s:%d", ip, port)
Beispiel #33
0
def main(host, port, bokeh_port, show, _bokeh, bokeh_whitelist, bokeh_prefix,
         use_xheaders, pid_file, scheduler_file, interface, local_directory,
         preload, preload_argv, tls_ca_file, tls_cert, tls_key):

    enable_proctitle_on_current()
    enable_proctitle_on_children()

    sec = Security(
        tls_ca_file=tls_ca_file,
        tls_scheduler_cert=tls_cert,
        tls_scheduler_key=tls_key,
    )

    if not host and (tls_ca_file or tls_cert or tls_key):
        host = 'tls://'

    if pid_file:
        with open(pid_file, 'w') as f:
            f.write(str(os.getpid()))

        def del_pid_file():
            if os.path.exists(pid_file):
                os.remove(pid_file)

        atexit.register(del_pid_file)

    local_directory_created = False
    if local_directory:
        if not os.path.exists(local_directory):
            os.mkdir(local_directory)
            local_directory_created = True
    else:
        local_directory = tempfile.mkdtemp(prefix='scheduler-')
        local_directory_created = True
    if local_directory not in sys.path:
        sys.path.insert(0, local_directory)

    if sys.platform.startswith('linux'):
        import resource  # module fails importing on Windows
        soft, hard = resource.getrlimit(resource.RLIMIT_NOFILE)
        limit = max(soft, hard // 2)
        resource.setrlimit(resource.RLIMIT_NOFILE, (limit, hard))

    if interface:
        if host:
            raise ValueError("Can not specify both interface and host")
        else:
            host = get_ip_interface(interface)

    addr = uri_from_host_port(host, port, 8786)

    loop = IOLoop.current()
    logger.info('-' * 47)

    services = {}
    if _bokeh:
        with ignoring(ImportError):
            from distributed.bokeh.scheduler import BokehScheduler
            services[('bokeh', bokeh_port)] = (BokehScheduler, {
                'prefix': bokeh_prefix
            })
    scheduler = Scheduler(loop=loop,
                          services=services,
                          scheduler_file=scheduler_file,
                          security=sec)
    scheduler.start(addr)
    preload_modules(preload,
                    parameter=scheduler,
                    file_dir=local_directory,
                    argv=preload_argv)

    logger.info('Local Directory: %26s', local_directory)
    logger.info('-' * 47)

    install_signal_handlers(loop)

    try:
        loop.start()
        loop.close()
    finally:
        scheduler.stop()
        if local_directory_created:
            shutil.rmtree(local_directory)

        logger.info("End scheduler at %r", addr)
Beispiel #34
0
def test_tls_scheduler(security, loop):
    s = Scheduler(security=security, loop=loop)
    s.start('localhost')
    assert s.address.startswith('tls')
    s.close()
Beispiel #35
0
def main(
    host,
    port,
    bokeh_port,
    show,
    dashboard,
    bokeh,
    dashboard_prefix,
    use_xheaders,
    pid_file,
    tls_ca_file,
    tls_cert,
    tls_key,
    dashboard_address,
    **kwargs
):
    g0, g1, g2 = gc.get_threshold()  # https://github.com/dask/distributed/issues/1653
    gc.set_threshold(g0 * 3, g1 * 3, g2 * 3)

    enable_proctitle_on_current()
    enable_proctitle_on_children()

    if bokeh_port is not None:
        warnings.warn(
            "The --bokeh-port flag has been renamed to --dashboard-address. "
            "Consider adding ``--dashboard-address :%d`` " % bokeh_port
        )
        dashboard_address = bokeh_port
    if bokeh is not None:
        warnings.warn(
            "The --bokeh/--no-bokeh flag has been renamed to --dashboard/--no-dashboard. "
        )
        dashboard = bokeh

    if port is None and (not host or not re.search(r":\d", host)):
        port = 8786

    sec = Security(
        **{
            k: v
            for k, v in [
                ("tls_ca_file", tls_ca_file),
                ("tls_scheduler_cert", tls_cert),
                ("tls_scheduler_key", tls_key),
            ]
            if v is not None
        }
    )

    if "DASK_INTERNAL_INHERIT_CONFIG" in os.environ:
        config = deserialize_for_cli(os.environ["DASK_INTERNAL_INHERIT_CONFIG"])
        # Update the global config given priority to the existing global config
        dask.config.update(dask.config.global_config, config, priority="old")

    if not host and (tls_ca_file or tls_cert or tls_key):
        host = "tls://"

    if pid_file:
        with open(pid_file, "w") as f:
            f.write(str(os.getpid()))

        def del_pid_file():
            if os.path.exists(pid_file):
                os.remove(pid_file)

        atexit.register(del_pid_file)

    if sys.platform.startswith("linux"):
        import resource  # module fails importing on Windows

        soft, hard = resource.getrlimit(resource.RLIMIT_NOFILE)
        limit = max(soft, hard // 2)
        resource.setrlimit(resource.RLIMIT_NOFILE, (limit, hard))

    loop = IOLoop.current()
    logger.info("-" * 47)

    scheduler = Scheduler(
        loop=loop,
        security=sec,
        host=host,
        port=port,
        dashboard_address=dashboard_address if dashboard else None,
        service_kwargs={"dashboard": {"prefix": dashboard_prefix}},
        **kwargs
    )
    logger.info("-" * 47)

    install_signal_handlers(loop)

    async def run():
        await scheduler
        await scheduler.finished()

    try:
        loop.run_sync(run)
    finally:
        scheduler.stop()

        logger.info("End scheduler at %r", scheduler.address)