X-Git-Url: https://git.arvados.org/arvados.git/blobdiff_plain/5dbf5c8ea2d9eb2bc8e10a03ca625f12ed71f12c..63b03a39adfd78961c5bbb6a3a2d02ccd8c92e4d:/sdk/python/arvados/keep.py diff --git a/sdk/python/arvados/keep.py b/sdk/python/arvados/keep.py index 262e68864d..8d0a89d777 100644 --- a/sdk/python/arvados/keep.py +++ b/sdk/python/arvados/keep.py @@ -58,6 +58,9 @@ class KeepLocator(object): self.permission_hint()] + self.hints if s is not None) + def stripped(self): + return "%s+%i" % (self.md5sum, self.size) + def _make_hex_prop(name, length): # Build and return a new property with the given name that # must be a hex string of the given length. @@ -171,8 +174,7 @@ class KeepBlockCache(object): def cap_cache(self): '''Cap the cache size to self.cache_max''' - self._cache_lock.acquire() - try: + with self._cache_lock: # Select all slots except those where ready.is_set() and content is # None (that means there was an error reading the block). self._cache = [c for c in self._cache if not (c.ready.is_set() and c.content is None)] @@ -183,30 +185,35 @@ class KeepBlockCache(object): del self._cache[i] break sm = sum([slot.size() for slot in self._cache]) - finally: - self._cache_lock.release() + + def _get(self, locator): + # Test if the locator is already in the cache + for i in xrange(0, len(self._cache)): + if self._cache[i].locator == locator: + n = self._cache[i] + if i != 0: + # move it to the front + del self._cache[i] + self._cache.insert(0, n) + return n + return None + + def get(self, locator): + with self._cache_lock: + return self._get(locator) def reserve_cache(self, locator): '''Reserve a cache slot for the specified locator, or return the existing slot.''' - self._cache_lock.acquire() - try: - # Test if the locator is already in the cache - for i in xrange(0, len(self._cache)): - if self._cache[i].locator == locator: - n = self._cache[i] - if i != 0: - # move it to the front - del self._cache[i] - self._cache.insert(0, n) - return n, False - - # Add a new cache slot for the locator - n = KeepBlockCache.CacheSlot(locator) - self._cache.insert(0, n) - return n, True - finally: - self._cache_lock.release() + with self._cache_lock: + n = self._get(locator) + if n: + return n, False + else: + # Add a new cache slot for the locator + n = KeepBlockCache.CacheSlot(locator) + self._cache.insert(0, n) + return n, True class KeepClient(object): @@ -279,10 +286,11 @@ class KeepClient(object): HTTP_ERRORS = (requests.exceptions.RequestException, socket.error, ssl.SSLError) - def __init__(self, root, **headers): + def __init__(self, root, session, **headers): self.root = root self.last_result = None self.success_flag = None + self.session = session self.get_headers = {'Accept': 'application/octet-stream'} self.get_headers.update(headers) self.put_headers = headers @@ -305,7 +313,7 @@ class KeepClient(object): _logger.debug("Request: GET %s", url) try: with timer.Timer() as t: - result = requests.get(url.encode('utf-8'), + result = self.session.get(url.encode('utf-8'), headers=self.get_headers, timeout=timeout) except self.HTTP_ERRORS as e: @@ -318,7 +326,7 @@ class KeepClient(object): content = result.content _logger.info("%s response: %s bytes in %s msec (%.3f MiB/sec)", self.last_status(), len(content), t.msecs, - (len(content)/(1024.0*1024))/t.secs) + (len(content)/(1024.0*1024))/t.secs if t.secs > 0 else 0) if self.success_flag: resp_md5 = hashlib.md5(content).hexdigest() if resp_md5 == locator.md5sum: @@ -331,7 +339,7 @@ class KeepClient(object): url = self.root + hash_s _logger.debug("Request: PUT %s", url) try: - result = requests.put(url.encode('utf-8'), + result = self.session.put(url.encode('utf-8'), data=body, headers=self.put_headers, timeout=timeout) @@ -371,9 +379,10 @@ class KeepClient(object): def run_with_limiter(self, limiter): if self.service.finished(): return - _logger.debug("KeepWriterThread %s proceeding %s %s", + _logger.debug("KeepWriterThread %s proceeding %s+%i %s", str(threading.current_thread()), self.args['data_hash'], + len(self.args['data']), self.args['service_root']) self._success = bool(self.service.put( self.args['data_hash'], @@ -382,9 +391,10 @@ class KeepClient(object): status = self.service.last_status() if self._success: result = self.service.last_result - _logger.debug("KeepWriterThread %s succeeded %s %s", + _logger.debug("KeepWriterThread %s succeeded %s+%i %s", str(threading.current_thread()), self.args['data_hash'], + len(self.args['data']), self.args['service_root']) # Tick the 'done' counter for the number of replica # reported stored by the server, for the case that @@ -394,49 +404,66 @@ class KeepClient(object): replicas_stored = int(result.headers['x-keep-replicas-stored']) except (KeyError, ValueError): replicas_stored = 1 - limiter.save_response(result.text.strip(), replicas_stored) + limiter.save_response(result.content.strip(), replicas_stored) elif status is not None: _logger.debug("Request fail: PUT %s => %s %s", self.args['data_hash'], status, - self.service.last_result.text) + self.service.last_result.content) def __init__(self, api_client=None, proxy=None, timeout=DEFAULT_TIMEOUT, proxy_timeout=DEFAULT_PROXY_TIMEOUT, api_token=None, local_store=None, block_cache=None, - num_retries=0): + num_retries=0, session=None): """Initialize a new KeepClient. Arguments: - * api_client: The API client to use to find Keep services. If not + :api_client: + The API client to use to find Keep services. If not provided, KeepClient will build one from available Arvados configuration. - * proxy: If specified, this KeepClient will send requests to this - Keep proxy. Otherwise, KeepClient will fall back to the setting - of the ARVADOS_KEEP_PROXY configuration setting. If you want to - ensure KeepClient does not use a proxy, pass in an empty string. - * timeout: The timeout (in seconds) for HTTP requests to Keep + + :proxy: + If specified, this KeepClient will send requests to this Keep + proxy. Otherwise, KeepClient will fall back to the setting of the + ARVADOS_KEEP_PROXY configuration setting. If you want to ensure + KeepClient does not use a proxy, pass in an empty string. + + :timeout: + The timeout (in seconds) for HTTP requests to Keep non-proxy servers. A tuple of two floats is interpreted as (connection_timeout, read_timeout): see http://docs.python-requests.org/en/latest/user/advanced/#timeouts. Default: (2, 300). - * proxy_timeout: The timeout (in seconds) for HTTP requests to + + :proxy_timeout: + The timeout (in seconds) for HTTP requests to Keep proxies. A tuple of two floats is interpreted as (connection_timeout, read_timeout). Default: (20, 300). - * api_token: If you're not using an API client, but only talking + + :api_token: + If you're not using an API client, but only talking directly to a Keep proxy, this parameter specifies an API token to authenticate Keep requests. It is an error to specify both api_client and api_token. If you specify neither, KeepClient will use one available from the Arvados configuration. - * local_store: If specified, this KeepClient will bypass Keep + + :local_store: + If specified, this KeepClient will bypass Keep services, and save data to the named directory. If unspecified, KeepClient will fall back to the setting of the $KEEP_LOCAL_STORE environment variable. If you want to ensure KeepClient does not use local storage, pass in an empty string. This is primarily intended to mock a server for testing. - * num_retries: The default number of times to retry failed requests. + + :num_retries: + The default number of times to retry failed requests. This will be used as the default num_retries value when get() and put() are called. Default 0. + + :session: + The requests.Session object to use for get() and put() requests. + Will create one if not specified. """ self.lock = threading.Lock() if proxy is None: @@ -462,6 +489,7 @@ class KeepClient(object): self.put = self.local_store_put else: self.num_retries = num_retries + self.session = session if session is not None else requests.Session() if proxy: if not proxy.endswith('/'): proxy += '/' @@ -553,7 +581,7 @@ class KeepClient(object): local_roots = self.weighted_service_roots(md5_s, force_rebuild) for root in local_roots: if root not in roots_map: - roots_map[root] = self.KeepService(root, **headers) + roots_map[root] = self.KeepService(root, self.session, **headers) return local_roots @staticmethod @@ -574,6 +602,14 @@ class KeepClient(object): else: return None + def get_from_cache(self, loc): + """Fetch a block only if is in the cache, otherwise return None.""" + slot = self.block_cache.get(loc) + if slot.ready.is_set(): + return slot.get() + else: + return None + @retry.retry_method def get(self, loc_s, num_retries=None): """Get data from Keep. @@ -599,7 +635,6 @@ class KeepClient(object): return ''.join(self.get(x) for x in loc_s.split(',')) locator = KeepLocator(loc_s) expect_hash = locator.md5sum - slot, first = self.block_cache.reserve_cache(expect_hash) if not first: v = slot.get() @@ -613,7 +648,7 @@ class KeepClient(object): hint_roots = ['http://keep.{}.arvadosapi.com/'.format(hint[2:]) for hint in locator.hints if hint.startswith('K@')] # Map root URLs their KeepService objects. - roots_map = {root: self.KeepService(root) for root in hint_roots} + roots_map = {root: self.KeepService(root, self.session) for root in hint_roots} blob = None loop = retry.RetryLoop(num_retries, self._check_loop_result, backoff_start=2) @@ -663,7 +698,7 @@ class KeepClient(object): "{} not found".format(loc_s), service_errors) else: raise arvados.errors.KeepReadError( - "failed to read {}".format(loc_s), service_errors) + "failed to read {}".format(loc_s), service_errors, label="service") @retry.retry_method def put(self, data, copies=2, num_retries=None): @@ -684,6 +719,12 @@ class KeepClient(object): exponential backoff. The default value is set when the KeepClient is initialized. """ + + if isinstance(data, unicode): + data = data.encode("ascii") + elif not isinstance(data, str): + raise arvados.errors.ArgumentError("Argument 'data' to KeepClient.put must be type 'str'") + data_hash = hashlib.md5(data).hexdigest() if copies < 1: return data_hash @@ -734,7 +775,7 @@ class KeepClient(object): if not roots_map[key].success_flag) raise arvados.errors.KeepWriteError( "failed to write {} (wanted {} copies but wrote {})".format( - data_hash, copies, thread_limiter.done()), service_errors) + data_hash, copies, thread_limiter.done()), service_errors, label="service") def local_store_put(self, data, copies=1, num_retries=None): """A stub for put(). @@ -767,3 +808,6 @@ class KeepClient(object): return '' with open(os.path.join(self.local_store, locator.md5sum), 'r') as f: return f.read() + + def is_cached(self, locator): + return self.block_cache.reserve_cache(expect_hash)