[Notes] [Git][BuildStream/buildstream][raoul/802-refactor-artifactcache] 6 commits: _cas: Rename artifactcache folder and move that to a root module

[Raoul Hidalgo Charman <gitlab mg gitlab com>]

Title: GitLab

Raoul Hidalgo Charman pushed to branch raoul/802-refactor-artifactcache at BuildStream / buildstream

Commits:

• c69d12f9
  by Raoul Hidalgo Charman at 2019-01-15T09:54:29Z

  _cas: Rename artifactcache folder and move that to a root module
  
  Other components will start to reply on cas modules, and not the artifact cache
  modules so it should be organized to reflect this.
  
  All relevant imports have been changed.
  
  Part #802
  

• fe01405b
  by Raoul Hidalgo Charman at 2019-01-15T09:54:34Z

  casremote.py: Move remote CAS classes into its own file
  
  Part of #802
  

• cdc5b6f5
  by Raoul Hidalgo Charman at 2019-01-15T09:54:47Z

  cas: move remote only functions to CASRemote
  
  List of methods moved
  * Initialization check: made it a class method that is run in a subprocess, for
    when checking in the main buildstream process.
  * fetch_blobs
  * send_blobs
  * verify_digest_on_remote
  * push_method
  
  Part of #802
  

• 38c5a533
  by Raoul Hidalgo Charman at 2019-01-15T09:54:47Z

  artifactcache: add top level tmpdir for CASRemote
  
  As CASRemote is used in other places such as SandboxRemote it makes sense for
  there to be a top level tmpdir it uses. This dir is passed to CASRemote when
  initialized.
  
  Part of #802
  

• e8861275
  by Raoul Hidalgo Charman at 2019-01-15T09:54:47Z

  artifactcache: Move pull logic into CASRemote
  
  Seperates the pull logic into a remote/local API, so that artifact cache
  iterates over blob digests checks whether it has them, and then requests them
  if not. The request command allows batching of blobs where appropriate.
  
  Tests have been updated to ensure the correct tmpdir is set up in process
  wrappers, else invalid cross link errors happen in the CI. Additional asserts
  have been added to check that the temporary directories are cleared by the end
  of a pull.
  
  Part of #802
  

• 89d9b6b8
  by Raoul Hidalgo Charman at 2019-01-15T09:54:47Z

  artifactcache: implement new push methods
  
  Similar to the pull methods, this implements a yield_directory_digests methods
  that iterates over blobs in the local CAS, with the upload_blob sending blobs
  to a remote and batching them where appropriate.
  
  Part of #802
  

22 changed files:

• buildstream/_artifactcache/artifactcache.py → buildstream/_artifactcache.py
• buildstream/_artifactcache/__init__.py → buildstream/_cas/__init__.py
• buildstream/_artifactcache/cascache.py → buildstream/_cas/cascache.py
• + buildstream/_cas/casremote.py
• buildstream/_artifactcache/casserver.py → buildstream/_cas/casserver.py
• + buildstream/_cas/transfer.py
• buildstream/_context.py
• buildstream/_exceptions.py
• buildstream/data/userconfig.yaml
• buildstream/sandbox/_sandboxremote.py
• conftest.py
• doc/source/using_configuring_artifact_server.rst
• tests/artifactcache/config.py
• tests/artifactcache/expiry.py
• tests/artifactcache/pull.py
• tests/artifactcache/push.py
• tests/integration/pullbuildtrees.py
• tests/sandboxes/storage-tests.py
• tests/storage/virtual_directory_import.py
• tests/testutils/artifactshare.py
• tests/testutils/runcli.py
• tests/utils/misc.py

Changes:

buildstream/_artifactcache/artifactcache.py → buildstream/_artifactcache.py

---

...	...	@@ -19,18 +19,17 @@
19	19
20	20	import multiprocessing
21	21	import os
22		-import signal
23	22	import string
24	23	from collections.abc import Mapping
25	24
26		-from ..types import _KeyStrength
27		-from .._exceptions import ArtifactError, CASError, LoadError, LoadErrorReason
28		-from .._message import Message, MessageType
29		-from .. import _signals
30		-from .. import utils
31		-from .. import _yaml
	25	+from .types import _KeyStrength
	26	+from ._exceptions import ArtifactError, CASError, LoadError, LoadErrorReason
	27	+from ._message import Message, MessageType
	28	+from . import utils
	29	+from . import _yaml
32	30
33		-from .cascache import CASRemote, CASRemoteSpec
	31	+from ._cas import BlobNotFound, CASRemote, CASRemoteSpec
	32	+from ._cas.transfer import cas_directory_upload, cas_directory_download, cas_tree_download
34	33
35	34
36	35	CACHE_SIZE_FILE = "cache_size"
...	...	@@ -375,20 +374,8 @@ class ArtifactCache():
375	374	remotes = {}
376	375	q = multiprocessing.Queue()
377	376	for remote_spec in remote_specs:
378		- # Use subprocess to avoid creation of gRPC threads in main BuildStream process
379		- # See https://github.com/grpc/grpc/blob/master/doc/fork_support.md for details
380		- p = multiprocessing.Process(target=self.cas.initialize_remote, args=(remote_spec, q))
381	377
382		- try:
383		- # Keep SIGINT blocked in the child process
384		- with _signals.blocked([signal.SIGINT], ignore=False):
385		- p.start()
386		-
387		- error = q.get()
388		- p.join()
389		- except KeyboardInterrupt:
390		- utils._kill_process_tree(p.pid)
391		- raise
	378	+ error = CASRemote.check_remote(remote_spec, self.context.tmpdir, q)
392	379
393	380	if error and on_failure:
394	381	on_failure(remote_spec.url, error)
...	...	@@ -399,7 +386,7 @@ class ArtifactCache():
399	386	if remote_spec.push:
400	387	self._has_push_remotes = True
401	388
402		- remotes[remote_spec.url] = CASRemote(remote_spec)
	389	+ remotes[remote_spec.url] = CASRemote(remote_spec, self.context.tmpdir)
403	390
404	391	for project in self.context.get_projects():
405	392	remote_specs = self.global_remote_specs
...	...	@@ -621,16 +608,41 @@ class ArtifactCache():
621	608
622	609	for remote in push_remotes:
623	610	remote.init()
	611	+ skipped_remote = True
624	612	display_key = element._get_brief_display_key()
625	613	element.status("Pushing artifact {} -> {}".format(display_key, remote.spec.url))
626	614
627		- if self.cas.push(refs, remote):
628		- element.info("Pushed artifact {} -> {}".format(display_key, remote.spec.url))
	615	+ try:
	616	+ for ref in refs:
	617	+ # Check whether ref is already on the server in which case
	618	+ # there is no need to push the ref
	619	+ root_digest = self.cas.resolve_ref(ref)
	620	+ response = remote.get_reference(ref)
	621	+ if (response is not None and
	622	+ response.hash == root_digest.hash and
	623	+ response.size_bytes == root_digest.size_bytes):
	624	+ element.info("Remote ({}) already has {} cached".format(
	625	+ remote.spec.url, element._get_brief_display_key()))
	626	+ continue
	627	+
	628	+ # upload blobs
	629	+ cas_directory_upload(self.cas, remote, root_digest)
	630	+ remote.update_reference(ref, root_digest)
	631	+
	632	+ skipped_remote = False
	633	+
	634	+ except CASError as e:
	635	+ if str(e.reason) == "StatusCode.RESOURCE_EXHAUSTED":
	636	+ element.warn("Failed to push element to {}: Resource exhuasted"
	637	+ .format(remote.spec.url))
	638	+ continue
	639	+ else:
	640	+ raise ArtifactError("Failed to push refs {}: {}".format(refs, e),
	641	+ temporary=True) from e
	642	+
	643	+ if skipped_remote is False:
629	644	pushed = True
630		- else:
631		- element.info("Remote ({}) already has {} cached".format(
632		- remote.spec.url, element._get_brief_display_key()
633		- ))
	645	+ element.info("Pushed artifact {} -> {}".format(display_key, remote.spec.url))
634	646
635	647	return pushed
636	648
...	...	@@ -658,19 +670,31 @@ class ArtifactCache():
658	670	display_key = element._get_brief_display_key()
659	671	element.status("Pulling artifact {} <- {}".format(display_key, remote.spec.url))
660	672
661		- if self.cas.pull(ref, remote, progress=progress, subdir=subdir, excluded_subdirs=excluded_subdirs):
662		- element.info("Pulled artifact {} <- {}".format(display_key, remote.spec.url))
663		- if subdir:
664		- # Attempt to extract subdir into artifact extract dir if it already exists
665		- # without containing the subdir. If the respective artifact extract dir does not
666		- # exist a complete extraction will complete.
667		- self.extract(element, key, subdir)
668		- # no need to pull from additional remotes
669		- return True
670		- else:
	673	+ root_digest = remote.get_reference(ref)
	674	+
	675	+ if not root_digest:
671	676	element.info("Remote ({}) does not have {} cached".format(
672		- remote.spec.url, element._get_brief_display_key()
673		- ))
	677	+ remote.spec.url, element._get_brief_display_key()))
	678	+ continue
	679	+
	680	+ try:
	681	+ cas_directory_download(self.cas, remote, root_digest, excluded_subdirs)
	682	+ except BlobNotFound:
	683	+ element.info("Remote ({}) is missing blobs for {}".format(
	684	+ remote.spec.url, element._get_brief_display_key()))
	685	+ continue
	686	+
	687	+ self.cas.set_ref(ref, root_digest)
	688	+
	689	+ if subdir:
	690	+ # Attempt to extract subdir into artifact extract dir if it already exists
	691	+ # without containing the subdir. If the respective artifact extract dir does not
	692	+ # exist a complete extraction will complete.
	693	+ self.extract(element, key, subdir)
	694	+
	695	+ element.info("Pulled artifact {} <- {}".format(display_key, remote.spec.url))
	696	+ # no need to pull from additional remotes
	697	+ return True
674	698
675	699	except CASError as e:
676	700	raise ArtifactError("Failed to pull artifact {}: {}".format(
...	...	@@ -685,15 +709,16 @@ class ArtifactCache():
685	709	#
686	710	# Args:
687	711	# project (Project): The current project
688		- # digest (Digest): The digest of the tree
	712	+ # tree_digest (Digest): The digest of the tree
689	713	#
690		- def pull_tree(self, project, digest):
	714	+ def pull_tree(self, project, tree_digest):
691	715	for remote in self._remotes[project]:
692		- digest = self.cas.pull_tree(remote, digest)
693		-
694		- if digest:
695		- # no need to pull from additional remotes
696		- return digest
	716	+ try:
	717	+ root_digest = cas_tree_download(self.cas, remote, tree_digest)
	718	+ except BlobNotFound:
	719	+ continue
	720	+ else:
	721	+ return root_digest
697	722
698	723	return None
699	724
...	...	@@ -722,7 +747,7 @@ class ArtifactCache():
722	747	return
723	748
724	749	for remote in push_remotes:
725		- self.cas.push_directory(remote, directory)
	750	+ cas_directory_upload(self.cas, remote, directory.ref)
726	751
727	752	# push_message():
728	753	#
...	...	@@ -747,7 +772,7 @@ class ArtifactCache():
747	772	"servers are configured as push remotes.")
748	773
749	774	for remote in push_remotes:
750		- message_digest = self.cas.push_message(remote, message)
	775	+ message_digest = remote.push_message(message)
751	776
752	777	return message_digest
753	778
...	...	@@ -931,6 +956,19 @@ class ArtifactCache():
931	956	stat = os.statvfs(volume)
932	957	return stat.f_bsize * stat.f_bavail, stat.f_bsize * stat.f_blocks
933	958
	959	+ def _fetch_directory(self, remote, root_digest, excluded_subdirs):
	960	+ for blob_digest in remote.yield_directory_digests(
	961	+ root_digest, excluded_subdirs=excluded_subdirs):
	962	+ if self.cas.check_blob(blob_digest):
	963	+ continue
	964	+ remote.request_blob(blob_digest)
	965	+ for blob_file in remote.get_blobs():
	966	+ self.cas.add_object(path=blob_file.name, link_directly=True)
	967	+
	968	+ # Request final CAS batch
	969	+ for blob_file in remote.get_blobs(complete_batch=True):
	970	+ self.cas.add_object(path=blob_file.name, link_directly=True)
	971	+
934	972
935	973	# _configured_remote_artifact_cache_specs():
936	974	#

buildstream/_artifactcache/__init__.py → buildstream/_cas/__init__.py

---

...	...	@@ -17,4 +17,5 @@
17	17	# Authors:
18	18	# Tristan Van Berkom <tristan vanberkom codethink co uk>
19	19
20		-from .artifactcache import ArtifactCache, ArtifactCacheSpec, CACHE_SIZE_FILE
	20	+from .cascache import CASCache
	21	+from .casremote import CASRemote, CASRemoteSpec, BlobNotFound

buildstream/_artifactcache/cascache.py → buildstream/_cas/cascache.py

---

...	...	@@ -17,85 +17,16 @@
17	17	# Authors:
18	18	# Jürg Billeter <juerg billeter codethink co uk>
19	19
20		-from collections import namedtuple
21	20	import hashlib
22		-import itertools
23		-import io
24	21	import os
25	22	import stat
26	23	import tempfile
27		-import uuid
28	24	import contextlib
29		-from urllib.parse import urlparse
30	25
31		-import grpc
32		-
33		-from .._protos.google.rpc import code_pb2
34		-from .._protos.google.bytestream import bytestream_pb2, bytestream_pb2_grpc
35		-from .._protos.build.bazel.remote.execution.v2 import remote_execution_pb2, remote_execution_pb2_grpc
36		-from .._protos.buildstream.v2 import buildstream_pb2, buildstream_pb2_grpc
	26	+from .._protos.build.bazel.remote.execution.v2 import remote_execution_pb2
37	27
38	28	from .. import utils
39		-from .._exceptions import CASError, LoadError, LoadErrorReason
40		-from .. import _yaml
41		-
42		-
43		-# The default limit for gRPC messages is 4 MiB.
44		-# Limit payload to 1 MiB to leave sufficient headroom for metadata.
45		-_MAX_PAYLOAD_BYTES = 1024 * 1024
46		-
47		-
48		-class CASRemoteSpec(namedtuple('CASRemoteSpec', 'url push server_cert client_key client_cert instance_name')):
49		-
50		- # _new_from_config_node
51		- #
52		- # Creates an CASRemoteSpec() from a YAML loaded node
53		- #
54		- @staticmethod
55		- def _new_from_config_node(spec_node, basedir=None):
56		- _yaml.node_validate(spec_node, ['url', 'push', 'server-cert', 'client-key', 'client-cert', 'instance-name'])
57		- url = _yaml.node_get(spec_node, str, 'url')
58		- push = _yaml.node_get(spec_node, bool, 'push', default_value=False)
59		- if not url:
60		- provenance = _yaml.node_get_provenance(spec_node, 'url')
61		- raise LoadError(LoadErrorReason.INVALID_DATA,
62		- "{}: empty artifact cache URL".format(provenance))
63		-
64		- instance_name = _yaml.node_get(spec_node, str, 'instance-name', default_value=None)
65		-
66		- server_cert = _yaml.node_get(spec_node, str, 'server-cert', default_value=None)
67		- if server_cert and basedir:
68		- server_cert = os.path.join(basedir, server_cert)
69		-
70		- client_key = _yaml.node_get(spec_node, str, 'client-key', default_value=None)
71		- if client_key and basedir:
72		- client_key = os.path.join(basedir, client_key)
73		-
74		- client_cert = _yaml.node_get(spec_node, str, 'client-cert', default_value=None)
75		- if client_cert and basedir:
76		- client_cert = os.path.join(basedir, client_cert)
77		-
78		- if client_key and not client_cert:
79		- provenance = _yaml.node_get_provenance(spec_node, 'client-key')
80		- raise LoadError(LoadErrorReason.INVALID_DATA,
81		- "{}: 'client-key' was specified without 'client-cert'".format(provenance))
82		-
83		- if client_cert and not client_key:
84		- provenance = _yaml.node_get_provenance(spec_node, 'client-cert')
85		- raise LoadError(LoadErrorReason.INVALID_DATA,
86		- "{}: 'client-cert' was specified without 'client-key'".format(provenance))
87		-
88		- return CASRemoteSpec(url, push, server_cert, client_key, client_cert, instance_name)
89		-
90		-
91		-CASRemoteSpec.__new__.__defaults__ = (None, None, None, None)
92		-
93		-
94		-class BlobNotFound(CASError):
95		-
96		- def __init__(self, blob, msg):
97		- self.blob = blob
98		- super().__init__(msg)
	29	+from .._exceptions import CASCacheError
99	30
100	31
101	32	# A CASCache manages a CAS repository as specified in the Remote Execution API.
...	...	@@ -120,7 +51,7 @@ class CASCache():
120	51	headdir = os.path.join(self.casdir, 'refs', 'heads')
121	52	objdir = os.path.join(self.casdir, 'objects')
122	53	if not (os.path.isdir(headdir) and os.path.isdir(objdir)):
123		- raise CASError("CAS repository check failed for '{}'".format(self.casdir))
	54	+ raise CASCacheError("CAS repository check failed for '{}'".format(self.casdir))
124	55
125	56	# contains():
126	57	#
...	...	@@ -169,7 +100,7 @@ class CASCache():
169	100	# subdir (str): Optional specific dir to extract
170	101	#
171	102	# Raises:
172		- # CASError: In cases there was an OSError, or if the ref did not exist.
	103	+ # CASCacheError: In cases there was an OSError, or if the ref did not exist.
173	104	#
174	105	# Returns: path to extracted directory
175	106	#
...	...	@@ -201,7 +132,7 @@ class CASCache():
201	132	# Another process beat us to rename
202	133	pass
203	134	except OSError as e:
204		- raise CASError("Failed to extract directory for ref '{}': {}".format(ref, e)) from e
	135	+ raise CASCacheError("Failed to extract directory for ref '{}': {}".format(ref, e)) from e
205	136
206	137	return originaldest
207	138
...	...	@@ -245,96 +176,6 @@ class CASCache():
245	176
246	177	return modified, removed, added
247	178
248		- def initialize_remote(self, remote_spec, q):
249		- try:
250		- remote = CASRemote(remote_spec)
251		- remote.init()
252		-
253		- request = buildstream_pb2.StatusRequest(instance_name=remote_spec.instance_name)
254		- response = remote.ref_storage.Status(request)
255		-
256		- if remote_spec.push and not response.allow_updates:
257		- q.put('CAS server does not allow push')
258		- else:
259		- # No error
260		- q.put(None)
261		-
262		- except grpc.RpcError as e:
263		- # str(e) is too verbose for errors reported to the user
264		- q.put(e.details())
265		-
266		- except Exception as e: # pylint: disable=broad-except
267		- # Whatever happens, we need to return it to the calling process
268		- #
269		- q.put(str(e))
270		-
271		- # pull():
272		- #
273		- # Pull a ref from a remote repository.
274		- #
275		- # Args:
276		- # ref (str): The ref to pull
277		- # remote (CASRemote): The remote repository to pull from
278		- # progress (callable): The progress callback, if any
279		- # subdir (str): The optional specific subdir to pull
280		- # excluded_subdirs (list): The optional list of subdirs to not pull
281		- #
282		- # Returns:
283		- # (bool): True if pull was successful, False if ref was not available
284		- #
285		- def pull(self, ref, remote, *, progress=None, subdir=None, excluded_subdirs=None):
286		- try:
287		- remote.init()
288		-
289		- request = buildstream_pb2.GetReferenceRequest(instance_name=remote.spec.instance_name)
290		- request.key = ref
291		- response = remote.ref_storage.GetReference(request)
292		-
293		- tree = remote_execution_pb2.Digest()
294		- tree.hash = response.digest.hash
295		- tree.size_bytes = response.digest.size_bytes
296		-
297		- # Check if the element artifact is present, if so just fetch the subdir.
298		- if subdir and os.path.exists(self.objpath(tree)):
299		- self._fetch_subdir(remote, tree, subdir)
300		- else:
301		- # Fetch artifact, excluded_subdirs determined in pullqueue
302		- self._fetch_directory(remote, tree, excluded_subdirs=excluded_subdirs)
303		-
304		- self.set_ref(ref, tree)
305		-
306		- return True
307		- except grpc.RpcError as e:
308		- if e.code() != grpc.StatusCode.NOT_FOUND:
309		- raise CASError("Failed to pull ref {}: {}".format(ref, e)) from e
310		- else:
311		- return False
312		- except BlobNotFound as e:
313		- return False
314		-
315		- # pull_tree():
316		- #
317		- # Pull a single Tree rather than a ref.
318		- # Does not update local refs.
319		- #
320		- # Args:
321		- # remote (CASRemote): The remote to pull from
322		- # digest (Digest): The digest of the tree
323		- #
324		- def pull_tree(self, remote, digest):
325		- try:
326		- remote.init()
327		-
328		- digest = self._fetch_tree(remote, digest)
329		-
330		- return digest
331		-
332		- except grpc.RpcError as e:
333		- if e.code() != grpc.StatusCode.NOT_FOUND:
334		- raise
335		-
336		- return None
337		-
338	179	# link_ref():
339	180	#
340	181	# Add an alias for an existing ref.
...	...	@@ -348,117 +189,6 @@ class CASCache():
348	189
349	190	self.set_ref(newref, tree)
350	191
351		- # push():
352		- #
353		- # Push committed refs to remote repository.
354		- #
355		- # Args:
356		- # refs (list): The refs to push
357		- # remote (CASRemote): The remote to push to
358		- #
359		- # Returns:
360		- # (bool): True if any remote was updated, False if no pushes were required
361		- #
362		- # Raises:
363		- # (CASError): if there was an error
364		- #
365		- def push(self, refs, remote):
366		- skipped_remote = True
367		- try:
368		- for ref in refs:
369		- tree = self.resolve_ref(ref)
370		-
371		- # Check whether ref is already on the server in which case
372		- # there is no need to push the ref
373		- try:
374		- request = buildstream_pb2.GetReferenceRequest(instance_name=remote.spec.instance_name)
375		- request.key = ref
376		- response = remote.ref_storage.GetReference(request)
377		-
378		- if response.digest.hash == tree.hash and response.digest.size_bytes == tree.size_bytes:
379		- # ref is already on the server with the same tree
380		- continue
381		-
382		- except grpc.RpcError as e:
383		- if e.code() != grpc.StatusCode.NOT_FOUND:
384		- # Intentionally re-raise RpcError for outer except block.
385		- raise
386		-
387		- self._send_directory(remote, tree)
388		-
389		- request = buildstream_pb2.UpdateReferenceRequest(instance_name=remote.spec.instance_name)
390		- request.keys.append(ref)
391		- request.digest.hash = tree.hash
392		- request.digest.size_bytes = tree.size_bytes
393		- remote.ref_storage.UpdateReference(request)
394		-
395		- skipped_remote = False
396		- except grpc.RpcError as e:
397		- if e.code() != grpc.StatusCode.RESOURCE_EXHAUSTED:
398		- raise CASError("Failed to push ref {}: {}".format(refs, e), temporary=True) from e
399		-
400		- return not skipped_remote
401		-
402		- # push_directory():
403		- #
404		- # Push the given virtual directory to a remote.
405		- #
406		- # Args:
407		- # remote (CASRemote): The remote to push to
408		- # directory (Directory): A virtual directory object to push.
409		- #
410		- # Raises:
411		- # (CASError): if there was an error
412		- #
413		- def push_directory(self, remote, directory):
414		- remote.init()
415		-
416		- self._send_directory(remote, directory.ref)
417		-
418		- # push_message():
419		- #
420		- # Push the given protobuf message to a remote.
421		- #
422		- # Args:
423		- # remote (CASRemote): The remote to push to
424		- # message (Message): A protobuf message to push.
425		- #
426		- # Raises:
427		- # (CASError): if there was an error
428		- #
429		- def push_message(self, remote, message):
430		-
431		- message_buffer = message.SerializeToString()
432		- message_digest = utils._message_digest(message_buffer)
433		-
434		- remote.init()
435		-
436		- with io.BytesIO(message_buffer) as b:
437		- self._send_blob(remote, message_digest, b)
438		-
439		- return message_digest
440		-
441		- # verify_digest_on_remote():
442		- #
443		- # Check whether the object is already on the server in which case
444		- # there is no need to upload it.
445		- #
446		- # Args:
447		- # remote (CASRemote): The remote to check
448		- # digest (Digest): The object digest.
449		- #
450		- def verify_digest_on_remote(self, remote, digest):
451		- remote.init()
452		-
453		- request = remote_execution_pb2.FindMissingBlobsRequest(instance_name=remote.spec.instance_name)
454		- request.blob_digests.extend([digest])
455		-
456		- response = remote.cas.FindMissingBlobs(request)
457		- if digest in response.missing_blob_digests:
458		- return False
459		-
460		- return True
461		-
462	192	# objpath():
463	193	#
464	194	# Return the path of an object based on its digest.
...	...	@@ -531,7 +261,7 @@ class CASCache():
531	261	pass
532	262
533	263	except OSError as e:
534		- raise CASError("Failed to hash object: {}".format(e)) from e
	264	+ raise CASCacheError("Failed to hash object: {}".format(e)) from e
535	265
536	266	return digest
537	267
...	...	@@ -572,7 +302,7 @@ class CASCache():
572	302	return digest
573	303
574	304	except FileNotFoundError as e:
575		- raise CASError("Attempt to access unavailable ref: {}".format(e)) from e
	305	+ raise CASCacheError("Attempt to access unavailable ref: {}".format(e)) from e
576	306
577	307	# update_mtime()
578	308	#
...	...	@@ -585,7 +315,7 @@ class CASCache():
585	315	try:
586	316	os.utime(self._refpath(ref))
587	317	except FileNotFoundError as e:
588		- raise CASError("Attempt to access unavailable ref: {}".format(e)) from e
	318	+ raise CASCacheError("Attempt to access unavailable ref: {}".format(e)) from e
589	319
590	320	# calculate_cache_size()
591	321	#
...	...	@@ -676,7 +406,7 @@ class CASCache():
676	406	# Remove cache ref
677	407	refpath = self._refpath(ref)
678	408	if not os.path.exists(refpath):
679		- raise CASError("Could not find ref '{}'".format(ref))
	409	+ raise CASCacheError("Could not find ref '{}'".format(ref))
680	410
681	411	os.unlink(refpath)
682	412
...	...	@@ -720,6 +450,37 @@ class CASCache():
720	450	reachable = set()
721	451	self._reachable_refs_dir(reachable, tree, update_mtime=True)
722	452
	453	+ # Check to see if a blob is in the local CAS
	454	+ # return None if not
	455	+ def check_blob(self, digest):
	456	+ objpath = self.objpath(digest)
	457	+ if os.path.exists(objpath):
	458	+ # already in local repository
	459	+ return objpath
	460	+ else:
	461	+ return None
	462	+
	463	+ def yield_directory_digests(self, directory_digest):
	464	+ # parse directory, and recursively add blobs
	465	+ d = remote_execution_pb2.Digest()
	466	+ d.hash = directory_digest.hash
	467	+ d.size_bytes = directory_digest.size_bytes
	468	+ yield d
	469	+
	470	+ directory = remote_execution_pb2.Directory()
	471	+
	472	+ with open(self.objpath(directory_digest), 'rb') as f:
	473	+ directory.ParseFromString(f.read())
	474	+
	475	+ for filenode in directory.files:
	476	+ d = remote_execution_pb2.Digest()
	477	+ d.hash = filenode.digest.hash
	478	+ d.size_bytes = filenode.digest.size_bytes
	479	+ yield d
	480	+
	481	+ for dirnode in directory.directories:
	482	+ yield from self.yield_directory_digests(dirnode.digest)
	483	+
723	484	################################################
724	485	# Local Private Methods #
725	486	################################################
...	...	@@ -792,7 +553,7 @@ class CASCache():
792	553	# The process serving the socket can't be cached anyway
793	554	pass
794	555	else:
795		- raise CASError("Unsupported file type for {}".format(full_path))
	556	+ raise CASCacheError("Unsupported file type for {}".format(full_path))
796	557
797	558	return self.add_object(digest=dir_digest,
798	559	buffer=directory.SerializeToString())
...	...	@@ -811,7 +572,7 @@ class CASCache():
811	572	if dirnode.name == name:
812	573	return dirnode.digest
813	574
814		- raise CASError("Subdirectory {} not found".format(name))
	575	+ raise CASCacheError("Subdirectory {} not found".format(name))
815	576
816	577	def _diff_trees(self, tree_a, tree_b, *, added, removed, modified, path=""):
817	578	dir_a = remote_execution_pb2.Directory()
...	...	@@ -908,429 +669,3 @@ class CASCache():
908	669
909	670	for dirnode in directory.directories:
910	671	yield from self._required_blobs(dirnode.digest)
911		-
912		- def _fetch_blob(self, remote, digest, stream):
913		- resource_name_components = ['blobs', digest.hash, str(digest.size_bytes)]
914		-
915		- if remote.spec.instance_name:
916		- resource_name_components.insert(0, remote.spec.instance_name)
917		-
918		- resource_name = '/'.join(resource_name_components)
919		-
920		- request = bytestream_pb2.ReadRequest()
921		- request.resource_name = resource_name
922		- request.read_offset = 0
923		- for response in remote.bytestream.Read(request):
924		- stream.write(response.data)
925		- stream.flush()
926		-
927		- assert digest.size_bytes == os.fstat(stream.fileno()).st_size
928		-
929		- # _ensure_blob():
930		- #
931		- # Fetch and add blob if it's not already local.
932		- #
933		- # Args:
934		- # remote (Remote): The remote to use.
935		- # digest (Digest): Digest object for the blob to fetch.
936		- #
937		- # Returns:
938		- # (str): The path of the object
939		- #
940		- def _ensure_blob(self, remote, digest):
941		- objpath = self.objpath(digest)
942		- if os.path.exists(objpath):
943		- # already in local repository
944		- return objpath
945		-
946		- with tempfile.NamedTemporaryFile(dir=self.tmpdir) as f:
947		- self._fetch_blob(remote, digest, f)
948		-
949		- added_digest = self.add_object(path=f.name, link_directly=True)
950		- assert added_digest.hash == digest.hash
951		-
952		- return objpath
953		-
954		- def _batch_download_complete(self, batch):
955		- for digest, data in batch.send():
956		- with tempfile.NamedTemporaryFile(dir=self.tmpdir) as f:
957		- f.write(data)
958		- f.flush()
959		-
960		- added_digest = self.add_object(path=f.name, link_directly=True)
961		- assert added_digest.hash == digest.hash
962		-
963		- # Helper function for _fetch_directory().
964		- def _fetch_directory_batch(self, remote, batch, fetch_queue, fetch_next_queue):
965		- self._batch_download_complete(batch)
966		-
967		- # All previously scheduled directories are now locally available,
968		- # move them to the processing queue.
969		- fetch_queue.extend(fetch_next_queue)
970		- fetch_next_queue.clear()
971		- return _CASBatchRead(remote)
972		-
973		- # Helper function for _fetch_directory().
974		- def _fetch_directory_node(self, remote, digest, batch, fetch_queue, fetch_next_queue, *, recursive=False):
975		- in_local_cache = os.path.exists(self.objpath(digest))
976		-
977		- if in_local_cache:
978		- # Skip download, already in local cache.
979		- pass
980		- elif (digest.size_bytes >= remote.max_batch_total_size_bytes or
981		- not remote.batch_read_supported):
982		- # Too large for batch request, download in independent request.
983		- self._ensure_blob(remote, digest)
984		- in_local_cache = True
985		- else:
986		- if not batch.add(digest):
987		- # Not enough space left in batch request.
988		- # Complete pending batch first.
989		- batch = self._fetch_directory_batch(remote, batch, fetch_queue, fetch_next_queue)
990		- batch.add(digest)
991		-
992		- if recursive:
993		- if in_local_cache:
994		- # Add directory to processing queue.
995		- fetch_queue.append(digest)
996		- else:
997		- # Directory will be available after completing pending batch.
998		- # Add directory to deferred processing queue.
999		- fetch_next_queue.append(digest)
1000		-
1001		- return batch
1002		-
1003		- # _fetch_directory():
1004		- #
1005		- # Fetches remote directory and adds it to content addressable store.
1006		- #
1007		- # Fetches files, symbolic links and recursively other directories in
1008		- # the remote directory and adds them to the content addressable
1009		- # store.
1010		- #
1011		- # Args:
1012		- # remote (Remote): The remote to use.
1013		- # dir_digest (Digest): Digest object for the directory to fetch.
1014		- # excluded_subdirs (list): The optional list of subdirs to not fetch
1015		- #
1016		- def _fetch_directory(self, remote, dir_digest, *, excluded_subdirs=None):
1017		- fetch_queue = [dir_digest]
1018		- fetch_next_queue = []
1019		- batch = _CASBatchRead(remote)
1020		- if not excluded_subdirs:
1021		- excluded_subdirs = []
1022		-
1023		- while len(fetch_queue) + len(fetch_next_queue) > 0:
1024		- if not fetch_queue:
1025		- batch = self._fetch_directory_batch(remote, batch, fetch_queue, fetch_next_queue)
1026		-
1027		- dir_digest = fetch_queue.pop(0)
1028		-
1029		- objpath = self._ensure_blob(remote, dir_digest)
1030		-
1031		- directory = remote_execution_pb2.Directory()
1032		- with open(objpath, 'rb') as f:
1033		- directory.ParseFromString(f.read())
1034		-
1035		- for dirnode in directory.directories:
1036		- if dirnode.name not in excluded_subdirs:
1037		- batch = self._fetch_directory_node(remote, dirnode.digest, batch,
1038		- fetch_queue, fetch_next_queue, recursive=True)
1039		-
1040		- for filenode in directory.files:
1041		- batch = self._fetch_directory_node(remote, filenode.digest, batch,
1042		- fetch_queue, fetch_next_queue)
1043		-
1044		- # Fetch final batch
1045		- self._fetch_directory_batch(remote, batch, fetch_queue, fetch_next_queue)
1046		-
1047		- def _fetch_subdir(self, remote, tree, subdir):
1048		- subdirdigest = self._get_subdir(tree, subdir)
1049		- self._fetch_directory(remote, subdirdigest)
1050		-
1051		- def _fetch_tree(self, remote, digest):
1052		- # download but do not store the Tree object
1053		- with tempfile.NamedTemporaryFile(dir=self.tmpdir) as out:
1054		- self._fetch_blob(remote, digest, out)
1055		-
1056		- tree = remote_execution_pb2.Tree()
1057		-
1058		- with open(out.name, 'rb') as f:
1059		- tree.ParseFromString(f.read())
1060		-
1061		- tree.children.extend([tree.root])
1062		- for directory in tree.children:
1063		- for filenode in directory.files:
1064		- self._ensure_blob(remote, filenode.digest)
1065		-
1066		- # place directory blob only in final location when we've downloaded
1067		- # all referenced blobs to avoid dangling references in the repository
1068		- dirbuffer = directory.SerializeToString()
1069		- dirdigest = self.add_object(buffer=dirbuffer)
1070		- assert dirdigest.size_bytes == len(dirbuffer)
1071		-
1072		- return dirdigest
1073		-
1074		- def _send_blob(self, remote, digest, stream, u_uid=uuid.uuid4()):
1075		- resource_name_components = ['uploads', str(u_uid), 'blobs',
1076		- digest.hash, str(digest.size_bytes)]
1077		-
1078		- if remote.spec.instance_name:
1079		- resource_name_components.insert(0, remote.spec.instance_name)
1080		-
1081		- resource_name = '/'.join(resource_name_components)
1082		-
1083		- def request_stream(resname, instream):
1084		- offset = 0
1085		- finished = False
1086		- remaining = digest.size_bytes
1087		- while not finished:
1088		- chunk_size = min(remaining, _MAX_PAYLOAD_BYTES)
1089		- remaining -= chunk_size
1090		-
1091		- request = bytestream_pb2.WriteRequest()
1092		- request.write_offset = offset
1093		- # max. _MAX_PAYLOAD_BYTES chunks
1094		- request.data = instream.read(chunk_size)
1095		- request.resource_name = resname
1096		- request.finish_write = remaining <= 0
1097		-
1098		- yield request
1099		-
1100		- offset += chunk_size
1101		- finished = request.finish_write
1102		-
1103		- response = remote.bytestream.Write(request_stream(resource_name, stream))
1104		-
1105		- assert response.committed_size == digest.size_bytes
1106		-
1107		- def _send_directory(self, remote, digest, u_uid=uuid.uuid4()):
1108		- required_blobs = self._required_blobs(digest)
1109		-
1110		- missing_blobs = dict()
1111		- # Limit size of FindMissingBlobs request
1112		- for required_blobs_group in _grouper(required_blobs, 512):
1113		- request = remote_execution_pb2.FindMissingBlobsRequest(instance_name=remote.spec.instance_name)
1114		-
1115		- for required_digest in required_blobs_group:
1116		- d = request.blob_digests.add()
1117		- d.hash = required_digest.hash
1118		- d.size_bytes = required_digest.size_bytes
1119		-
1120		- response = remote.cas.FindMissingBlobs(request)
1121		- for missing_digest in response.missing_blob_digests:
1122		- d = remote_execution_pb2.Digest()
1123		- d.hash = missing_digest.hash
1124		- d.size_bytes = missing_digest.size_bytes
1125		- missing_blobs[d.hash] = d
1126		-
1127		- # Upload any blobs missing on the server
1128		- self._send_blobs(remote, missing_blobs.values(), u_uid)
1129		-
1130		- def _send_blobs(self, remote, digests, u_uid=uuid.uuid4()):
1131		- batch = _CASBatchUpdate(remote)
1132		-
1133		- for digest in digests:
1134		- with open(self.objpath(digest), 'rb') as f:
1135		- assert os.fstat(f.fileno()).st_size == digest.size_bytes
1136		-
1137		- if (digest.size_bytes >= remote.max_batch_total_size_bytes or
1138		- not remote.batch_update_supported):
1139		- # Too large for batch request, upload in independent request.
1140		- self._send_blob(remote, digest, f, u_uid=u_uid)
1141		- else:
1142		- if not batch.add(digest, f):
1143		- # Not enough space left in batch request.
1144		- # Complete pending batch first.
1145		- batch.send()
1146		- batch = _CASBatchUpdate(remote)
1147		- batch.add(digest, f)
1148		-
1149		- # Send final batch
1150		- batch.send()
1151		-
1152		-
1153		-# Represents a single remote CAS cache.
1154		-#
1155		-class CASRemote():
1156		- def __init__(self, spec):
1157		- self.spec = spec
1158		- self._initialized = False
1159		- self.channel = None
1160		- self.bytestream = None
1161		- self.cas = None
1162		- self.ref_storage = None
1163		- self.batch_update_supported = None
1164		- self.batch_read_supported = None
1165		- self.capabilities = None
1166		- self.max_batch_total_size_bytes = None
1167		-
1168		- def init(self):
1169		- if not self._initialized:
1170		- url = urlparse(self.spec.url)
1171		- if url.scheme == 'http':
1172		- port = url.port or 80
1173		- self.channel = grpc.insecure_channel('{}:{}'.format(url.hostname, port))
1174		- elif url.scheme == 'https':
1175		- port = url.port or 443
1176		-
1177		- if self.spec.server_cert:
1178		- with open(self.spec.server_cert, 'rb') as f:
1179		- server_cert_bytes = f.read()
1180		- else:
1181		- server_cert_bytes = None
1182		-
1183		- if self.spec.client_key:
1184		- with open(self.spec.client_key, 'rb') as f:
1185		- client_key_bytes = f.read()
1186		- else:
1187		- client_key_bytes = None
1188		-
1189		- if self.spec.client_cert:
1190		- with open(self.spec.client_cert, 'rb') as f:
1191		- client_cert_bytes = f.read()
1192		- else:
1193		- client_cert_bytes = None
1194		-
1195		- credentials = grpc.ssl_channel_credentials(root_certificates=server_cert_bytes,
1196		- private_key=client_key_bytes,
1197		- certificate_chain=client_cert_bytes)
1198		- self.channel = grpc.secure_channel('{}:{}'.format(url.hostname, port), credentials)
1199		- else:
1200		- raise CASError("Unsupported URL: {}".format(self.spec.url))
1201		-
1202		- self.bytestream = bytestream_pb2_grpc.ByteStreamStub(self.channel)
1203		- self.cas = remote_execution_pb2_grpc.ContentAddressableStorageStub(self.channel)
1204		- self.capabilities = remote_execution_pb2_grpc.CapabilitiesStub(self.channel)
1205		- self.ref_storage = buildstream_pb2_grpc.ReferenceStorageStub(self.channel)
1206		-
1207		- self.max_batch_total_size_bytes = _MAX_PAYLOAD_BYTES
1208		- try:
1209		- request = remote_execution_pb2.GetCapabilitiesRequest(instance_name=self.spec.instance_name)
1210		- response = self.capabilities.GetCapabilities(request)
1211		- server_max_batch_total_size_bytes = response.cache_capabilities.max_batch_total_size_bytes
1212		- if 0 < server_max_batch_total_size_bytes < self.max_batch_total_size_bytes:
1213		- self.max_batch_total_size_bytes = server_max_batch_total_size_bytes
1214		- except grpc.RpcError as e:
1215		- # Simply use the defaults for servers that don't implement GetCapabilities()
1216		- if e.code() != grpc.StatusCode.UNIMPLEMENTED:
1217		- raise
1218		-
1219		- # Check whether the server supports BatchReadBlobs()
1220		- self.batch_read_supported = False
1221		- try:
1222		- request = remote_execution_pb2.BatchReadBlobsRequest(instance_name=self.spec.instance_name)
1223		- response = self.cas.BatchReadBlobs(request)
1224		- self.batch_read_supported = True
1225		- except grpc.RpcError as e:
1226		- if e.code() != grpc.StatusCode.UNIMPLEMENTED:
1227		- raise
1228		-
1229		- # Check whether the server supports BatchUpdateBlobs()
1230		- self.batch_update_supported = False
1231		- try:
1232		- request = remote_execution_pb2.BatchUpdateBlobsRequest(instance_name=self.spec.instance_name)
1233		- response = self.cas.BatchUpdateBlobs(request)
1234		- self.batch_update_supported = True
1235		- except grpc.RpcError as e:
1236		- if (e.code() != grpc.StatusCode.UNIMPLEMENTED and
1237		- e.code() != grpc.StatusCode.PERMISSION_DENIED):
1238		- raise
1239		-
1240		- self._initialized = True
1241		-
1242		-
1243		-# Represents a batch of blobs queued for fetching.
1244		-#
1245		-class _CASBatchRead():
1246		- def __init__(self, remote):
1247		- self._remote = remote
1248		- self._max_total_size_bytes = remote.max_batch_total_size_bytes
1249		- self._request = remote_execution_pb2.BatchReadBlobsRequest(instance_name=remote.spec.instance_name)
1250		- self._size = 0
1251		- self._sent = False
1252		-
1253		- def add(self, digest):
1254		- assert not self._sent
1255		-
1256		- new_batch_size = self._size + digest.size_bytes
1257		- if new_batch_size > self._max_total_size_bytes:
1258		- # Not enough space left in current batch
1259		- return False
1260		-
1261		- request_digest = self._request.digests.add()
1262		- request_digest.hash = digest.hash
1263		- request_digest.size_bytes = digest.size_bytes
1264		- self._size = new_batch_size
1265		- return True
1266		-
1267		- def send(self):
1268		- assert not self._sent
1269		- self._sent = True
1270		-
1271		- if not self._request.digests:
1272		- return
1273		-
1274		- batch_response = self._remote.cas.BatchReadBlobs(self._request)
1275		-
1276		- for response in batch_response.responses:
1277		- if response.status.code == code_pb2.NOT_FOUND:
1278		- raise BlobNotFound(response.digest.hash, "Failed to download blob {}: {}".format(
1279		- response.digest.hash, response.status.code))
1280		- if response.status.code != code_pb2.OK:
1281		- raise CASError("Failed to download blob {}: {}".format(
1282		- response.digest.hash, response.status.code))
1283		- if response.digest.size_bytes != len(response.data):
1284		- raise CASError("Failed to download blob {}: expected {} bytes, received {} bytes".format(
1285		- response.digest.hash, response.digest.size_bytes, len(response.data)))
1286		-
1287		- yield (response.digest, response.data)
1288		-
1289		-
1290		-# Represents a batch of blobs queued for upload.
1291		-#
1292		-class _CASBatchUpdate():
1293		- def __init__(self, remote):
1294		- self._remote = remote
1295		- self._max_total_size_bytes = remote.max_batch_total_size_bytes
1296		- self._request = remote_execution_pb2.BatchUpdateBlobsRequest(instance_name=remote.spec.instance_name)
1297		- self._size = 0
1298		- self._sent = False
1299		-
1300		- def add(self, digest, stream):
1301		- assert not self._sent
1302		-
1303		- new_batch_size = self._size + digest.size_bytes
1304		- if new_batch_size > self._max_total_size_bytes:
1305		- # Not enough space left in current batch
1306		- return False
1307		-
1308		- blob_request = self._request.requests.add()
1309		- blob_request.digest.hash = digest.hash
1310		- blob_request.digest.size_bytes = digest.size_bytes
1311		- blob_request.data = stream.read(digest.size_bytes)
1312		- self._size = new_batch_size
1313		- return True
1314		-
1315		- def send(self):
1316		- assert not self._sent
1317		- self._sent = True
1318		-
1319		- if not self._request.requests:
1320		- return
1321		-
1322		- batch_response = self._remote.cas.BatchUpdateBlobs(self._request)
1323		-
1324		- for response in batch_response.responses:
1325		- if response.status.code != code_pb2.OK:
1326		- raise CASError("Failed to upload blob {}: {}".format(
1327		- response.digest.hash, response.status.code))
1328		-
1329		-
1330		-def _grouper(iterable, n):
1331		- while True:
1332		- try:
1333		- current = next(iterable)
1334		- except StopIteration:
1335		- return
1336		- yield itertools.chain([current], itertools.islice(iterable, n - 1))

buildstream/_cas/casremote.py

---

	1	+from collections import namedtuple
	2	+import io
	3	+import itertools
	4	+import os
	5	+import multiprocessing
	6	+import signal
	7	+import tempfile
	8	+from urllib.parse import urlparse
	9	+import uuid
	10	+
	11	+import grpc
	12	+
	13	+from .. import _yaml
	14	+from .._protos.google.rpc import code_pb2
	15	+from .._protos.google.bytestream import bytestream_pb2, bytestream_pb2_grpc
	16	+from .._protos.build.bazel.remote.execution.v2 import remote_execution_pb2, remote_execution_pb2_grpc
	17	+from .._protos.buildstream.v2 import buildstream_pb2, buildstream_pb2_grpc
	18	+
	19	+from .._exceptions import CASRemoteError, LoadError, LoadErrorReason
	20	+from .. import _signals
	21	+from .. import utils
	22	+
	23	+# The default limit for gRPC messages is 4 MiB.
	24	+# Limit payload to 1 MiB to leave sufficient headroom for metadata.
	25	+_MAX_PAYLOAD_BYTES = 1024 * 1024
	26	+
	27	+
	28	+class CASRemoteSpec(namedtuple('CASRemoteSpec', 'url push server_cert client_key client_cert instance_name')):
	29	+
	30	+ # _new_from_config_node
	31	+ #
	32	+ # Creates an CASRemoteSpec() from a YAML loaded node
	33	+ #
	34	+ @staticmethod
	35	+ def _new_from_config_node(spec_node, basedir=None):
	36	+ _yaml.node_validate(spec_node, ['url', 'push', 'server-cert', 'client-key', 'client-cert', 'instance_name'])
	37	+ url = _yaml.node_get(spec_node, str, 'url')
	38	+ push = _yaml.node_get(spec_node, bool, 'push', default_value=False)
	39	+ if not url:
	40	+ provenance = _yaml.node_get_provenance(spec_node, 'url')
	41	+ raise LoadError(LoadErrorReason.INVALID_DATA,
	42	+ "{}: empty artifact cache URL".format(provenance))
	43	+
	44	+ instance_name = _yaml.node_get(spec_node, str, 'server-cert', default_value=None)
	45	+
	46	+ server_cert = _yaml.node_get(spec_node, str, 'server-cert', default_value=None)
	47	+ if server_cert and basedir:
	48	+ server_cert = os.path.join(basedir, server_cert)
	49	+
	50	+ client_key = _yaml.node_get(spec_node, str, 'client-key', default_value=None)
	51	+ if client_key and basedir:
	52	+ client_key = os.path.join(basedir, client_key)
	53	+
	54	+ client_cert = _yaml.node_get(spec_node, str, 'client-cert', default_value=None)
	55	+ if client_cert and basedir:
	56	+ client_cert = os.path.join(basedir, client_cert)
	57	+
	58	+ if client_key and not client_cert:
	59	+ provenance = _yaml.node_get_provenance(spec_node, 'client-key')
	60	+ raise LoadError(LoadErrorReason.INVALID_DATA,
	61	+ "{}: 'client-key' was specified without 'client-cert'".format(provenance))
	62	+
	63	+ if client_cert and not client_key:
	64	+ provenance = _yaml.node_get_provenance(spec_node, 'client-cert')
	65	+ raise LoadError(LoadErrorReason.INVALID_DATA,
	66	+ "{}: 'client-cert' was specified without 'client-key'".format(provenance))
	67	+
	68	+ return CASRemoteSpec(url, push, server_cert, client_key, client_cert, instance_name)
	69	+
	70	+
	71	+CASRemoteSpec.__new__.__defaults__ = (None, None, None, None)
	72	+
	73	+
	74	+class BlobNotFound(CASRemoteError):
	75	+
	76	+ def __init__(self, blob, msg):
	77	+ self.blob = blob
	78	+ super().__init__(msg)
	79	+
	80	+
	81	+# Represents a single remote CAS cache.
	82	+#
	83	+class CASRemote():
	84	+ def __init__(self, spec, tmpdir):
	85	+ self.spec = spec
	86	+ self._initialized = False
	87	+ self.channel = None
	88	+ self.bytestream = None
	89	+ self.cas = None
	90	+ self.ref_storage = None
	91	+ self.batch_update_supported = None
	92	+ self.batch_read_supported = None
	93	+ self.capabilities = None
	94	+ self.max_batch_total_size_bytes = None
	95	+
	96	+ # Need str because python 3.5 and lower doesn't deal with path like
	97	+ # objects here.
	98	+ self.tmpdir = str(tmpdir)
	99	+ os.makedirs(self.tmpdir, exist_ok=True)
	100	+
	101	+ self.__tmp_downloads = [] # files in the tmpdir waiting to be added to local caches
	102	+
	103	+ self.__batch_read = None
	104	+ self.__batch_update = None
	105	+
	106	+ def init(self):
	107	+ if not self._initialized:
	108	+ url = urlparse(self.spec.url)
	109	+ if url.scheme == 'http':
	110	+ port = url.port or 80
	111	+ self.channel = grpc.insecure_channel('{}:{}'.format(url.hostname, port))
	112	+ elif url.scheme == 'https':
	113	+ port = url.port or 443
	114	+
	115	+ if self.spec.server_cert:
	116	+ with open(self.spec.server_cert, 'rb') as f:
	117	+ server_cert_bytes = f.read()
	118	+ else:
	119	+ server_cert_bytes = None
	120	+
	121	+ if self.spec.client_key:
	122	+ with open(self.spec.client_key, 'rb') as f:
	123	+ client_key_bytes = f.read()
	124	+ else:
	125	+ client_key_bytes = None
	126	+
	127	+ if self.spec.client_cert:
	128	+ with open(self.spec.client_cert, 'rb') as f:
	129	+ client_cert_bytes = f.read()
	130	+ else:
	131	+ client_cert_bytes = None
	132	+
	133	+ credentials = grpc.ssl_channel_credentials(root_certificates=server_cert_bytes,
	134	+ private_key=client_key_bytes,
	135	+ certificate_chain=client_cert_bytes)
	136	+ self.channel = grpc.secure_channel('{}:{}'.format(url.hostname, port), credentials)
	137	+ else:
	138	+ raise CASRemoteError("Unsupported URL: {}".format(self.spec.url))
	139	+
	140	+ self.bytestream = bytestream_pb2_grpc.ByteStreamStub(self.channel)
	141	+ self.cas = remote_execution_pb2_grpc.ContentAddressableStorageStub(self.channel)
	142	+ self.capabilities = remote_execution_pb2_grpc.CapabilitiesStub(self.channel)
	143	+ self.ref_storage = buildstream_pb2_grpc.ReferenceStorageStub(self.channel)
	144	+
	145	+ self.max_batch_total_size_bytes = _MAX_PAYLOAD_BYTES
	146	+ try:
	147	+ request = remote_execution_pb2.GetCapabilitiesRequest()
	148	+ response = self.capabilities.GetCapabilities(request)
	149	+ server_max_batch_total_size_bytes = response.cache_capabilities.max_batch_total_size_bytes
	150	+ if 0 < server_max_batch_total_size_bytes < self.max_batch_total_size_bytes:
	151	+ self.max_batch_total_size_bytes = server_max_batch_total_size_bytes
	152	+ except grpc.RpcError as e:
	153	+ # Simply use the defaults for servers that don't implement GetCapabilities()
	154	+ if e.code() != grpc.StatusCode.UNIMPLEMENTED:
	155	+ raise
	156	+
	157	+ # Check whether the server supports BatchReadBlobs()
	158	+ self.batch_read_supported = False
	159	+ try:
	160	+ request = remote_execution_pb2.BatchReadBlobsRequest()
	161	+ response = self.cas.BatchReadBlobs(request)
	162	+ self.batch_read_supported = True
	163	+ self.__batch_read = _CASBatchRead(self)
	164	+ except grpc.RpcError as e:
	165	+ if e.code() != grpc.StatusCode.UNIMPLEMENTED:
	166	+ raise
	167	+
	168	+ # Check whether the server supports BatchUpdateBlobs()
	169	+ self.batch_update_supported = False
	170	+ try:
	171	+ request = remote_execution_pb2.BatchUpdateBlobsRequest()
	172	+ response = self.cas.BatchUpdateBlobs(request)
	173	+ self.batch_update_supported = True
	174	+ self.__batch_update = _CASBatchUpdate(self)
	175	+ except grpc.RpcError as e:
	176	+ if (e.code() != grpc.StatusCode.UNIMPLEMENTED and
	177	+ e.code() != grpc.StatusCode.PERMISSION_DENIED):
	178	+ raise
	179	+
	180	+ self._initialized = True
	181	+
	182	+ # check_remote
	183	+ #
	184	+ # Used when checking whether remote_specs work in the buildstream main
	185	+ # thread, runs this in a seperate process to avoid creation of gRPC threads
	186	+ # in the main BuildStream process
	187	+ # See https://github.com/grpc/grpc/blob/master/doc/fork_support.md for details
	188	+ @classmethod
	189	+ def check_remote(cls, remote_spec, tmpdir, q):
	190	+
	191	+ def __check_remote():
	192	+ try:
	193	+ remote = cls(remote_spec, tmpdir)
	194	+ remote.init()
	195	+
	196	+ request = buildstream_pb2.StatusRequest()
	197	+ response = remote.ref_storage.Status(request)
	198	+
	199	+ if remote_spec.push and not response.allow_updates:
	200	+ q.put('CAS server does not allow push')
	201	+ else:
	202	+ # No error
	203	+ q.put(None)
	204	+
	205	+ except grpc.RpcError as e:
	206	+ # str(e) is too verbose for errors reported to the user
	207	+ q.put(e.details())
	208	+
	209	+ except Exception as e: # pylint: disable=broad-except
	210	+ # Whatever happens, we need to return it to the calling process
	211	+ #
	212	+ q.put(str(e))
	213	+
	214	+ p = multiprocessing.Process(target=__check_remote)
	215	+
	216	+ try:
	217	+ # Keep SIGINT blocked in the child process
	218	+ with _signals.blocked([signal.SIGINT], ignore=False):
	219	+ p.start()
	220	+
	221	+ error = q.get()
	222	+ p.join()
	223	+ except KeyboardInterrupt:
	224	+ utils._kill_process_tree(p.pid)
	225	+ raise
	226	+
	227	+ return error
	228	+
	229	+ # verify_digest_on_remote():
	230	+ #
	231	+ # Check whether the object is already on the server in which case
	232	+ # there is no need to upload it.
	233	+ #
	234	+ # Args:
	235	+ # digest (Digest): The object digest.
	236	+ #
	237	+ def verify_digest_on_remote(self, digest):
	238	+ self.init()
	239	+
	240	+ request = remote_execution_pb2.FindMissingBlobsRequest()
	241	+ request.blob_digests.extend([digest])
	242	+
	243	+ response = self.cas.FindMissingBlobs(request)
	244	+ if digest in response.missing_blob_digests:
	245	+ return False
	246	+
	247	+ return True
	248	+
	249	+ # push_message():
	250	+ #
	251	+ # Push the given protobuf message to a remote.
	252	+ #
	253	+ # Args:
	254	+ # message (Message): A protobuf message to push.
	255	+ #
	256	+ # Raises:
	257	+ # (CASRemoteError): if there was an error
	258	+ #
	259	+ def push_message(self, message):
	260	+
	261	+ message_buffer = message.SerializeToString()
	262	+ message_digest = utils._message_digest(message_buffer)
	263	+
	264	+ self.init()
	265	+
	266	+ with io.BytesIO(message_buffer) as b:
	267	+ self._send_blob(message_digest, b)
	268	+
	269	+ return message_digest
	270	+
	271	+ # get_reference():
	272	+ #
	273	+ # Args:
	274	+ # ref (str): The ref to request
	275	+ #
	276	+ # Returns:
	277	+ # (digest): digest of ref, None if not found
	278	+ #
	279	+ def get_reference(self, ref):
	280	+ try:
	281	+ self.init()
	282	+
	283	+ request = buildstream_pb2.GetReferenceRequest()
	284	+ request.key = ref
	285	+ return self.ref_storage.GetReference(request).digest
	286	+ except grpc.RpcError as e:
	287	+ if e.code() != grpc.StatusCode.NOT_FOUND:
	288	+ raise CASRemoteError("Failed to find ref {}: {}".format(ref, e)) from e
	289	+ else:
	290	+ return None
	291	+
	292	+ # update_reference():
	293	+ #
	294	+ # Args:
	295	+ # ref (str): Reference to update
	296	+ # digest (Digest): New digest to update ref with
	297	+ def update_reference(self, ref, digest):
	298	+ request = buildstream_pb2.UpdateReferenceRequest()
	299	+ request.keys.append(ref)
	300	+ request.digest.hash = digest.hash
	301	+ request.digest.size_bytes = digest.size_bytes
	302	+ self.ref_storage.UpdateReference(request)
	303	+
	304	+ def get_tree_blob(self, tree_digest):
	305	+ self.init()
	306	+ f = tempfile.NamedTemporaryFile(dir=self.tmpdir)
	307	+ self._fetch_blob(tree_digest, f)
	308	+
	309	+ tree = remote_execution_pb2.Tree()
	310	+ with open(f.name, 'rb') as tmp:
	311	+ tree.ParseFromString(tmp.read())
	312	+
	313	+ return tree
	314	+
	315	+ # yield_directory_digests():
	316	+ #
	317	+ # Recursively iterates over digests for files, symbolic links and other
	318	+ # directories starting from a root digest
	319	+ #
	320	+ # Args:
	321	+ # root_digest (digest): The root_digest to get a tree of
	322	+ # progress (callable): The progress callback, if any
	323	+ # subdir (str): The optional specific subdir to pull
	324	+ # excluded_subdirs (list): The optional list of subdirs to not pull
	325	+ #
	326	+ # Returns:
	327	+ # (iter digests): recursively iterates over digests contained in root directory
	328	+ #
	329	+ def yield_directory_digests(self, root_digest, *, progress=None,
	330	+ subdir=None, excluded_subdirs=None):
	331	+ self.init()
	332	+
	333	+ # Fetch artifact, excluded_subdirs determined in pullqueue
	334	+ if excluded_subdirs is None:
	335	+ excluded_subdirs = []
	336	+
	337	+ # get directory blob
	338	+ f = tempfile.NamedTemporaryFile(dir=self.tmpdir)
	339	+ self._fetch_blob(root_digest, f)
	340	+
	341	+ directory = remote_execution_pb2.Directory()
	342	+ with open(f.name, 'rb') as tmp:
	343	+ directory.ParseFromString(tmp.read())
	344	+
	345	+ yield root_digest
	346	+ for filenode in directory.files:
	347	+ yield filenode.digest
	348	+
	349	+ for dirnode in directory.directories:
	350	+ if dirnode.name not in excluded_subdirs:
	351	+ yield from self.yield_directory_digests(dirnode.digest)
	352	+
	353	+ # yield_tree_digests():
	354	+ #
	355	+ # Fetches a tree file from digests and then iterates over child digests
	356	+ #
	357	+ # Args:
	358	+ # tree_digest (digest): tree digest
	359	+ #
	360	+ # Returns:
	361	+ # (iter digests): iterates over digests in tree message
	362	+ def yield_tree_digests(self, tree):
	363	+ self.init()
	364	+
	365	+ tree.children.extend([tree.root])
	366	+ for directory in tree.children:
	367	+ for filenode in directory.files:
	368	+ yield filenode.digest
	369	+
	370	+ # add the directory to downloaded tmp files to be added
	371	+ f = tempfile.NamedTemporaryFile(dir=self.tmpdir)
	372	+ f.write(directory.SerializeToString())
	373	+ f.flush()
	374	+ self.__tmp_downloads.append(f)
	375	+
	376	+ # request_blob():
	377	+ #
	378	+ # Request blob, triggering download depending via bytestream or cas
	379	+ # BatchReadBlobs depending on size.
	380	+ #
	381	+ # Args:
	382	+ # digest (Digest): digest of the requested blob
	383	+ #
	384	+ def request_blob(self, digest):
	385	+ if (not self.batch_read_supported or
	386	+ digest.size_bytes > self.max_batch_total_size_bytes):
	387	+ f = tempfile.NamedTemporaryFile(dir=self.tmpdir)
	388	+ self._fetch_blob(digest, f)
	389	+ self.__tmp_downloads.append(f)
	390	+ elif self.__batch_read.add(digest) is False:
	391	+ self._download_batch()
	392	+ self.__batch_read.add(digest)
	393	+
	394	+ # get_blobs():
	395	+ #
	396	+ # Yield over downloaded blobs in the tmp file locations, causing the files
	397	+ # to be deleted once they go out of scope.
	398	+ #
	399	+ # Args:
	400	+ # complete_batch (bool): download any outstanding batch read request
	401	+ #
	402	+ # Returns:
	403	+ # iterator over NamedTemporaryFile
	404	+ def get_blobs(self, complete_batch=False):
	405	+ # Send read batch request and download
	406	+ if (complete_batch is True and
	407	+ self.batch_read_supported is True):
	408	+ self._download_batch()
	409	+
	410	+ while self.__tmp_downloads:
	411	+ yield self.__tmp_downloads.pop()
	412	+
	413	+ # upload_blob():
	414	+ #
	415	+ # Push blobs given an iterator over blob files
	416	+ #
	417	+ # Args:
	418	+ # digest (Digest): digest we want to upload
	419	+ # blob_file (str): Name of file location
	420	+ # u_uid (str): Used to identify to the bytestream service
	421	+ #
	422	+ def upload_blob(self, digest, blob_file, u_uid=uuid.uuid4()):
	423	+ with open(blob_file, 'rb') as f:
	424	+ assert os.fstat(f.fileno()).st_size == digest.size_bytes
	425	+
	426	+ if (digest.size_bytes >= self.max_batch_total_size_bytes or
	427	+ not self.batch_update_supported):
	428	+ # Too large for batch request, upload in independent request.
	429	+ self._send_blob(digest, f, u_uid=u_uid)
	430	+ else:
	431	+ if self.__batch_update.add(digest, f) is False:
	432	+ self.__batch_update.send()
	433	+ self.__batch_update = _CASBatchUpdate(self)
	434	+ self.__batch_update.add(digest, f)
	435	+
	436	+ # send_update_batch():
	437	+ #
	438	+ # Sends anything left in the update batch
	439	+ #
	440	+ def send_update_batch(self):
	441	+ # make sure everything is sent
	442	+ self.__batch_update.send()
	443	+ self.__batch_update = _CASBatchUpdate(self)
	444	+
	445	+ # find_missing_blobs()
	446	+ #
	447	+ # Does FindMissingBlobs request to remote
	448	+ #
	449	+ # Args:
	450	+ # required_blobs ([Digest]): list of blobs required
	451	+ #
	452	+ # Returns:
	453	+ # (Dict(Digest)): missing blobs
	454	+ def find_missing_blobs(self, required_blobs):
	455	+ self.init()
	456	+ missing_blobs = dict()
	457	+ # Limit size of FindMissingBlobs request
	458	+ for required_blobs_group in _grouper(required_blobs, 512):
	459	+ request = remote_execution_pb2.FindMissingBlobsRequest()
	460	+
	461	+ for required_digest in required_blobs_group:
	462	+ d = request.blob_digests.add()
	463	+ d.hash = required_digest.hash
	464	+ d.size_bytes = required_digest.size_bytes
	465	+
	466	+ response = self.cas.FindMissingBlobs(request)
	467	+ for missing_digest in response.missing_blob_digests:
	468	+ d = remote_execution_pb2.Digest()
	469	+ d.hash = missing_digest.hash
	470	+ d.size_bytes = missing_digest.size_bytes
	471	+ missing_blobs[d.hash] = d
	472	+
	473	+ return missing_blobs
	474	+
	475	+ ################################################
	476	+ # Local Private Methods #
	477	+ ################################################
	478	+ def _fetch_blob(self, digest, stream):
	479	+ resource_name = '/'.join(['blobs', digest.hash, str(digest.size_bytes)])
	480	+ request = bytestream_pb2.ReadRequest()
	481	+ request.resource_name = resource_name
	482	+ request.read_offset = 0
	483	+ for response in self.bytestream.Read(request):
	484	+ stream.write(response.data)
	485	+ stream.flush()
	486	+
	487	+ assert digest.size_bytes == os.fstat(stream.fileno()).st_size
	488	+
	489	+ def _send_blob(self, digest, stream, u_uid=uuid.uuid4()):
	490	+ resource_name = '/'.join(['uploads', str(u_uid), 'blobs',
	491	+ digest.hash, str(digest.size_bytes)])
	492	+
	493	+ def request_stream(resname, instream):
	494	+ offset = 0
	495	+ finished = False
	496	+ remaining = digest.size_bytes
	497	+ while not finished:
	498	+ chunk_size = min(remaining, _MAX_PAYLOAD_BYTES)
	499	+ remaining -= chunk_size
	500	+
	501	+ request = bytestream_pb2.WriteRequest()
	502	+ request.write_offset = offset
	503	+ # max. _MAX_PAYLOAD_BYTES chunks
	504	+ request.data = instream.read(chunk_size)
	505	+ request.resource_name = resname
	506	+ request.finish_write = remaining <= 0
	507	+
	508	+ yield request
	509	+
	510	+ offset += chunk_size
	511	+ finished = request.finish_write
	512	+
	513	+ try:
	514	+ response = self.bytestream.Write(request_stream(resource_name, stream))
	515	+ except grpc.RpcError as e:
	516	+ raise CASRemoteError("Failed to upload blob: {}".format(e), reason=e.code())
	517	+
	518	+ assert response.committed_size == digest.size_bytes
	519	+
	520	+ def _download_batch(self):
	521	+ for _, data in self.__batch_read.send():
	522	+ f = tempfile.NamedTemporaryFile(dir=self.tmpdir)
	523	+ f.write(data)
	524	+ f.flush()
	525	+ self.__tmp_downloads.append(f)
	526	+
	527	+ self.__batch_read = _CASBatchRead(self)
	528	+
	529	+
	530	+def _grouper(iterable, n):
	531	+ while True:
	532	+ try:
	533	+ current = next(iterable)
	534	+ except StopIteration:
	535	+ return
	536	+ yield itertools.chain([current], itertools.islice(iterable, n - 1))
	537	+
	538	+
	539	+# Represents a batch of blobs queued for fetching.
	540	+#
	541	+class _CASBatchRead():
	542	+ def __init__(self, remote):
	543	+ self._remote = remote
	544	+ self._max_total_size_bytes = remote.max_batch_total_size_bytes
	545	+ self._request = remote_execution_pb2.BatchReadBlobsRequest()
	546	+ self._size = 0
	547	+ self._sent = False
	548	+
	549	+ def add(self, digest):
	550	+ assert not self._sent
	551	+
	552	+ new_batch_size = self._size + digest.size_bytes
	553	+ if new_batch_size > self._max_total_size_bytes:
	554	+ # Not enough space left in current batch
	555	+ return False
	556	+
	557	+ request_digest = self._request.digests.add()
	558	+ request_digest.hash = digest.hash
	559	+ request_digest.size_bytes = digest.size_bytes
	560	+ self._size = new_batch_size
	561	+ return True
	562	+
	563	+ def send(self):
	564	+ assert not self._sent
	565	+ self._sent = True
	566	+
	567	+ if not self._request.digests:
	568	+ return
	569	+
	570	+ try:
	571	+ batch_response = self._remote.cas.BatchReadBlobs(self._request)
	572	+ except grpc.RpcError as e:
	573	+ raise CASRemoteError("Failed to read blob batch: {}".format(e),
	574	+ reason=e.code()) from e
	575	+
	576	+ for response in batch_response.responses:
	577	+ if response.status.code == code_pb2.NOT_FOUND:
	578	+ raise BlobNotFound(response.digest.hash, "Failed to download blob {}: {}".format(
	579	+ response.digest.hash, response.status.code))
	580	+ if response.status.code != code_pb2.OK:
	581	+ raise CASRemoteError("Failed to download blob {}: {}".format(
	582	+ response.digest.hash, response.status.code))
	583	+ if response.digest.size_bytes != len(response.data):
	584	+ raise CASRemoteError("Failed to download blob {}: expected {} bytes, received {} bytes".format(
	585	+ response.digest.hash, response.digest.size_bytes, len(response.data)))
	586	+
	587	+ yield (response.digest, response.data)
	588	+
	589	+
	590	+# Represents a batch of blobs queued for upload.
	591	+#
	592	+class _CASBatchUpdate():
	593	+ def __init__(self, remote):
	594	+ self._remote = remote
	595	+ self._max_total_size_bytes = remote.max_batch_total_size_bytes
	596	+ self._request = remote_execution_pb2.BatchUpdateBlobsRequest()
	597	+ self._size = 0
	598	+ self._sent = False
	599	+
	600	+ def add(self, digest, stream):
	601	+ assert not self._sent
	602	+
	603	+ new_batch_size = self._size + digest.size_bytes
	604	+ if new_batch_size > self._max_total_size_bytes:
	605	+ # Not enough space left in current batch
	606	+ return False
	607	+
	608	+ blob_request = self._request.requests.add()
	609	+ blob_request.digest.hash = digest.hash
	610	+ blob_request.digest.size_bytes = digest.size_bytes
	611	+ blob_request.data = stream.read(digest.size_bytes)
	612	+ self._size = new_batch_size
	613	+ return True
	614	+
	615	+ def send(self):
	616	+ assert not self._sent
	617	+ self._sent = True
	618	+
	619	+ if not self._request.requests:
	620	+ return
	621	+
	622	+ # Want to raise a CASRemoteError if
	623	+ try:
	624	+ batch_response = self._remote.cas.BatchUpdateBlobs(self._request)
	625	+ except grpc.RpcError as e:
	626	+ raise CASRemoteError("Failed to upload blob batch: {}".format(e),
	627	+ reason=e.code()) from e
	628	+
	629	+ for response in batch_response.responses:
	630	+ if response.status.code != code_pb2.OK:
	631	+ raise CASRemoteError("Failed to upload blob {}: {}".format(
	632	+ response.digest.hash, response.status.code))

buildstream/_artifactcache/casserver.py → buildstream/_cas/casserver.py

---

buildstream/_cas/transfer.py

---

	1	+#
	2	+# Copyright (C) 2017-2018 Codethink Limited
	3	+#
	4	+# This program is free software; you can redistribute it and/or
	5	+# modify it under the terms of the GNU Lesser General Public
	6	+# License as published by the Free Software Foundation; either
	7	+# version 2 of the License, or (at your option) any later version.
	8	+#
	9	+# This library is distributed in the hope that it will be useful,
	10	+# but WITHOUT ANY WARRANTY; without even the implied warranty of
	11	+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
	12	+# Lesser General Public License for more details.
	13	+#
	14	+# You should have received a copy of the GNU Lesser General Public
	15	+# License along with this library. If not, see <http://www.gnu.org/licenses/>.
	16	+#
	17	+# Authors:
	18	+# Raoul Hidalgo Charman <raoul hidalgocharman codethink co uk>
	19	+
	20	+from ..utils import _message_digest
	21	+
	22	+
	23	+def cas_directory_download(caslocal, casremote, root_digest, excluded_subdirs):
	24	+ for blob_digest in casremote.yield_directory_digests(
	25	+ root_digest, excluded_subdirs=excluded_subdirs):
	26	+ if caslocal.check_blob(blob_digest):
	27	+ continue
	28	+ casremote.request_blob(blob_digest)
	29	+ for blob_file in casremote.get_blobs():
	30	+ caslocal.add_object(path=blob_file.name, link_directly=True)
	31	+
	32	+ # Request final CAS batch
	33	+ for blob_file in casremote.get_blobs(complete_batch=True):
	34	+ caslocal.add_object(path=blob_file.name, link_directly=True)
	35	+
	36	+
	37	+def cas_tree_download(caslocal, casremote, tree_digest):
	38	+ tree = casremote.get_tree_blob(tree_digest)
	39	+ for blob_digest in casremote.yield_tree_digests(tree):
	40	+ if caslocal.check_blob(blob_digest):
	41	+ continue
	42	+ casremote.request_blob(blob_digest)
	43	+ for blob_file in casremote.get_blobs():
	44	+ caslocal.add_object(path=blob_file.name, link_directly=True)
	45	+
	46	+ # Get the last batch
	47	+ for blob_file in casremote.get_blobs(complete_batch=True):
	48	+ caslocal.add_object(path=blob_file.name, link_directly=True)
	49	+
	50	+ # get root digest from tree and return that
	51	+ return _message_digest(tree.root.SerializeToString())
	52	+
	53	+
	54	+def cas_directory_upload(caslocal, casremote, root_digest):
	55	+ required_blobs = caslocal.yield_directory_digests(root_digest)
	56	+ missing_blobs = casremote.find_missing_blobs(required_blobs)
	57	+ for blob in missing_blobs.values():
	58	+ blob_file = caslocal.objpath(blob)
	59	+ casremote.upload_blob(blob, blob_file)
	60	+
	61	+ # send remaining blobs
	62	+ casremote.send_update_batch()

buildstream/_context.py

---

...	...	@@ -31,7 +31,7 @@ from ._exceptions import LoadError, LoadErrorReason, BstError
31	31	from ._message import Message, MessageType
32	32	from ._profile import Topics, profile_start, profile_end
33	33	from ._artifactcache import ArtifactCache
34		-from ._artifactcache.cascache import CASCache
	34	+from ._cas import CASCache
35	35	from ._workspaces import Workspaces, WorkspaceProjectCache, WORKSPACE_PROJECT_FILE
36	36	from .plugin import _plugin_lookup
37	37	from .sandbox import SandboxRemote
...	...	@@ -191,10 +191,11 @@ class Context():
191	191	_yaml.node_validate(defaults, [
192	192	'sourcedir', 'builddir', 'artifactdir', 'logdir',
193	193	'scheduler', 'artifacts', 'logging', 'projects',
194		- 'cache', 'prompt', 'workspacedir', 'remote-execution'
	194	+ 'cache', 'prompt', 'workspacedir', 'remote-execution', 'tmpdir',
195	195	])
196	196
197		- for directory in ['sourcedir', 'builddir', 'artifactdir', 'logdir', 'workspacedir']:
	197	+ for directory in ['sourcedir', 'builddir', 'artifactdir', 'logdir',
	198	+ 'tmpdir', 'workspacedir']:
198	199	# Allow the ~ tilde expansion and any environment variables in
199	200	# path specification in the config files.
200	201	#

buildstream/_exceptions.py

---

...	...	@@ -284,6 +284,21 @@ class CASError(BstError):
284	284	super().__init__(message, detail=detail, domain=ErrorDomain.CAS, reason=reason, temporary=True)
285	285
286	286
	287	+# CASRemoteError
	288	+#
	289	+# Raised when errors are encountered in the remote CAS
	290	+class CASRemoteError(CASError):
	291	+ pass
	292	+
	293	+
	294	+# CASCacheError
	295	+#
	296	+# Raised when errors are encountered in the local CASCacheError
	297	+#
	298	+class CASCacheError(CASError):
	299	+ pass
	300	+
	301	+
287	302	# PipelineError
288	303	#
289	304	# Raised from pipeline operations

buildstream/data/userconfig.yaml

---

...	...	@@ -19,6 +19,9 @@ builddir: ${XDG_CACHE_HOME}/buildstream/build
19	19	# Location to store local binary artifacts
20	20	artifactdir: ${XDG_CACHE_HOME}/buildstream/artifacts
21	21
	22	+# Location to store temporary files, e.g. used when downloading from a casremote
	23	+tmpdir: ${XDG_CACHE_HOME}/buildstream/tmp
	24	+
22	25	# Location to store build logs
23	26	logdir: ${XDG_CACHE_HOME}/buildstream/logs
24	27

buildstream/sandbox/_sandboxremote.py

---

...	...	@@ -38,7 +38,8 @@ from .._protos.google.rpc import code_pb2
38	38	from .._exceptions import SandboxError
39	39	from .. import _yaml
40	40	from .._protos.google.longrunning import operations_pb2, operations_pb2_grpc
41		-from .._artifactcache.cascache import CASRemote, CASRemoteSpec
	41	+from .._cas import CASRemote, CASRemoteSpec
	42	+from .._cas.transfer import cas_tree_download, cas_directory_upload
42	43
43	44
44	45	class RemoteExecutionSpec(namedtuple('RemoteExecutionSpec', 'exec_service storage_service action_service')):
...	...	@@ -279,10 +280,9 @@ class SandboxRemote(Sandbox):
279	280
280	281	context = self._get_context()
281	282	cascache = context.get_cascache()
282		- casremote = CASRemote(self.storage_remote_spec)
	283	+ casremote = CASRemote(self.storage_remote_spec, context.tmpdir)
283	284
284		- # Now do a pull to ensure we have the necessary parts.
285		- dir_digest = cascache.pull_tree(casremote, tree_digest)
	285	+ dir_digest = cas_tree_download(cascache, casremote, tree_digest)
286	286	if dir_digest is None or not dir_digest.hash or not dir_digest.size_bytes:
287	287	raise SandboxError("Output directory structure pulling from remote failed.")
288	288
...	...	@@ -306,8 +306,9 @@ class SandboxRemote(Sandbox):
306	306
307	307	def _run(self, command, flags, *, cwd, env):
308	308	# set up virtual dircetory
	309	+ context = self._get_context()
309	310	upload_vdir = self.get_virtual_directory()
310		- cascache = self._get_context().get_cascache()
	311	+ cascache = context.get_cascache()
311	312	if isinstance(upload_vdir, FileBasedDirectory):
312	313	# Make a new temporary directory to put source in
313	314	upload_vdir = CasBasedDirectory(cascache, ref=None)
...	...	@@ -340,25 +341,25 @@ class SandboxRemote(Sandbox):
340	341	action_result = self._check_action_cache(action_digest)
341	342
342	343	if not action_result:
343		- casremote = CASRemote(self.storage_remote_spec)
	344	+ casremote = CASRemote(self.storage_remote_spec, context.tmpdir)
344	345
345	346	# Now, push that key (without necessarily needing a ref) to the remote.
346	347	try:
347		- cascache.push_directory(casremote, upload_vdir)
	348	+ cas_directory_upload(cascache, casremote, upload_vdir.ref)
348	349	except grpc.RpcError as e:
349	350	raise SandboxError("Failed to push source directory to remote: {}".format(e)) from e
350	351
351		- if not cascache.verify_digest_on_remote(casremote, upload_vdir.ref):
	352	+ if not casremote.verify_digest_on_remote(upload_vdir.ref):
352	353	raise SandboxError("Failed to verify that source has been pushed to the remote artifact cache.")
353	354
354	355	# Push command and action
355	356	try:
356		- cascache.push_message(casremote, command_proto)
	357	+ casremote.push_message(command_proto)
357	358	except grpc.RpcError as e:
358	359	raise SandboxError("Failed to push command to remote: {}".format(e))
359	360
360	361	try:
361		- cascache.push_message(casremote, action)
	362	+ casremote.push_message(action)
362	363	except grpc.RpcError as e:
363	364	raise SandboxError("Failed to push action to remote: {}".format(e))
364	365

conftest.py

---

...	...	@@ -46,6 +46,13 @@ def integration_cache(request):
46	46	else:
47	47	cache_dir = os.path.abspath('./integration-cache')
48	48
	49	+ # Clean up the tmp dir, should be empty but something in CI tests is
	50	+ # leaving files here
	51	+ try:
	52	+ shutil.rmtree(os.path.join(cache_dir, 'tmp'))
	53	+ except FileNotFoundError:
	54	+ pass
	55	+
49	56	yield cache_dir
50	57
51	58	# Clean up the artifacts after each test run - we only want to

doc/source/using_configuring_artifact_server.rst

---

...	...	@@ -94,7 +94,7 @@ requiring BuildStream's more exigent dependencies by setting the
94	94	Command reference
95	95	~~~~~~~~~~~~~~~~~
96	96
97		-.. click:: buildstream._artifactcache.casserver:server_main
	97	+.. click:: buildstream._cas.casserver:server_main
98	98	:prog: bst-artifact-server
99	99
100	100

tests/artifactcache/config.py

---

...	...	@@ -3,8 +3,7 @@ import pytest
3	3	import itertools
4	4	import os
5	5
6		-from buildstream._artifactcache import ArtifactCacheSpec
7		-from buildstream._artifactcache.artifactcache import _configured_remote_artifact_cache_specs
	6	+from buildstream._artifactcache import ArtifactCacheSpec, _configured_remote_artifact_cache_specs
8	7	from buildstream._context import Context
9	8	from buildstream._project import Project
10	9	from buildstream.utils import _deduplicate

tests/artifactcache/expiry.py

---

...	...	@@ -342,13 +342,13 @@ def test_invalid_cache_quota(cli, datafiles, tmpdir, quota, success):
342	342	total_space = 10000
343	343
344	344	volume_space_patch = mock.patch(
345		- "buildstream._artifactcache.artifactcache.ArtifactCache._get_volume_space_info_for",
	345	+ "buildstream._artifactcache.ArtifactCache._get_volume_space_info_for",
346	346	autospec=True,
347	347	return_value=(free_space, total_space),
348	348	)
349	349
350	350	cache_size_patch = mock.patch(
351		- "buildstream._artifactcache.artifactcache.ArtifactCache.get_cache_size",
	351	+ "buildstream._artifactcache.ArtifactCache.get_cache_size",
352	352	autospec=True,
353	353	return_value=0,
354	354	)

tests/artifactcache/pull.py

---

...	...	@@ -110,7 +110,7 @@ def test_pull(cli, tmpdir, datafiles):
110	110	# See https://github.com/grpc/grpc/blob/master/doc/fork_support.md for details
111	111	process = multiprocessing.Process(target=_queue_wrapper,
112	112	args=(_test_pull, queue, user_config_file, project_dir,
113		- artifact_dir, 'target.bst', element_key))
	113	+ artifact_dir, tmpdir, 'target.bst', element_key))
114	114
115	115	try:
116	116	# Keep SIGINT blocked in the child process
...	...	@@ -126,14 +126,18 @@ def test_pull(cli, tmpdir, datafiles):
126	126	assert not error
127	127	assert cas.contains(element, element_key)
128	128
	129	+ # Check that the tmp dir is cleared out
	130	+ assert os.listdir(os.path.join(str(tmpdir), 'cache', 'tmp')) == []
129	131
130		-def _test_pull(user_config_file, project_dir, artifact_dir,
	132	+
	133	+def _test_pull(user_config_file, project_dir, artifact_dir, tmpdir,
131	134	element_name, element_key, queue):
132	135	# Fake minimal context
133	136	context = Context()
134	137	context.load(config=user_config_file)
135	138	context.artifactdir = artifact_dir
136	139	context.set_message_handler(message_handler)
	140	+ context.tmpdir = os.path.join(str(tmpdir), 'cache', 'tmp')
137	141
138	142	# Load the project manually
139	143	project = Project(project_dir, context)
...	...	@@ -218,7 +222,7 @@ def test_pull_tree(cli, tmpdir, datafiles):
218	222	# See https://github.com/grpc/grpc/blob/master/doc/fork_support.md for details
219	223	process = multiprocessing.Process(target=_queue_wrapper,
220	224	args=(_test_push_tree, queue, user_config_file, project_dir,
221		- artifact_dir, artifact_digest))
	225	+ artifact_dir, tmpdir, artifact_digest))
222	226
223	227	try:
224	228	# Keep SIGINT blocked in the child process
...	...	@@ -239,6 +243,9 @@ def test_pull_tree(cli, tmpdir, datafiles):
239	243	# Assert that we are not cached locally anymore
240	244	assert cli.get_element_state(project_dir, 'target.bst') != 'cached'
241	245
	246	+ # Check that the tmp dir is cleared out
	247	+ assert os.listdir(os.path.join(str(tmpdir), 'cache', 'tmp')) == []
	248	+
242	249	tree_digest = remote_execution_pb2.Digest(hash=tree_hash,
243	250	size_bytes=tree_size)
244	251
...	...	@@ -246,7 +253,7 @@ def test_pull_tree(cli, tmpdir, datafiles):
246	253	# Use subprocess to avoid creation of gRPC threads in main BuildStream process
247	254	process = multiprocessing.Process(target=_queue_wrapper,
248	255	args=(_test_pull_tree, queue, user_config_file, project_dir,
249		- artifact_dir, tree_digest))
	256	+ artifact_dir, tmpdir, tree_digest))
250	257
251	258	try:
252	259	# Keep SIGINT blocked in the child process
...	...	@@ -267,13 +274,18 @@ def test_pull_tree(cli, tmpdir, datafiles):
267	274	# Ensure the entire Tree stucture has been pulled
268	275	assert os.path.exists(cas.objpath(directory_digest))
269	276
	277	+ # Check that the tmp dir is cleared out
	278	+ assert os.listdir(os.path.join(str(tmpdir), 'cache', 'tmp')) == []
	279	+
270	280
271		-def _test_push_tree(user_config_file, project_dir, artifact_dir, artifact_digest, queue):
	281	+def _test_push_tree(user_config_file, project_dir, artifact_dir, tmpdir,
	282	+ artifact_digest, queue):
272	283	# Fake minimal context
273	284	context = Context()
274	285	context.load(config=user_config_file)
275	286	context.artifactdir = artifact_dir
276	287	context.set_message_handler(message_handler)
	288	+ context.tmpdir = os.path.join(str(tmpdir), 'cache', 'tmp')
277	289
278	290	# Load the project manually
279	291	project = Project(project_dir, context)
...	...	@@ -304,12 +316,14 @@ def _test_push_tree(user_config_file, project_dir, artifact_dir, artifact_digest
304	316	queue.put("No remote configured")
305	317
306	318
307		-def _test_pull_tree(user_config_file, project_dir, artifact_dir, artifact_digest, queue):
	319	+def _test_pull_tree(user_config_file, project_dir, artifact_dir, tmpdir,
	320	+ artifact_digest, queue):
308	321	# Fake minimal context
309	322	context = Context()
310	323	context.load(config=user_config_file)
311	324	context.artifactdir = artifact_dir
312	325	context.set_message_handler(message_handler)
	326	+ context.tmpdir = os.path.join(str(tmpdir), 'cache', 'tmp')
313	327
314	328	# Load the project manually
315	329	project = Project(project_dir, context)

tests/artifactcache/push.py

---

...	...	@@ -89,7 +89,7 @@ def test_push(cli, tmpdir, datafiles):
89	89	# See https://github.com/grpc/grpc/blob/master/doc/fork_support.md for details
90	90	process = multiprocessing.Process(target=_queue_wrapper,
91	91	args=(_test_push, queue, user_config_file, project_dir,
92		- artifact_dir, 'target.bst', element_key))
	92	+ artifact_dir, tmpdir, 'target.bst', element_key))
93	93
94	94	try:
95	95	# Keep SIGINT blocked in the child process
...	...	@@ -105,14 +105,18 @@ def test_push(cli, tmpdir, datafiles):
105	105	assert not error
106	106	assert share.has_artifact('test', 'target.bst', element_key)
107	107
	108	+ # Check tmpdir for downloads is cleared
	109	+ assert os.listdir(os.path.join(str(tmpdir), 'cache', 'tmp')) == []
108	110
109		-def _test_push(user_config_file, project_dir, artifact_dir,
	111	+
	112	+def _test_push(user_config_file, project_dir, artifact_dir, tmpdir,
110	113	element_name, element_key, queue):
111	114	# Fake minimal context
112	115	context = Context()
113	116	context.load(config=user_config_file)
114	117	context.artifactdir = artifact_dir
115	118	context.set_message_handler(message_handler)
	119	+ context.tmpdir = os.path.join(str(tmpdir), 'cache', 'tmp')
116	120
117	121	# Load the project manually
118	122	project = Project(project_dir, context)
...	...	@@ -196,9 +200,10 @@ def test_push_directory(cli, tmpdir, datafiles):
196	200	queue = multiprocessing.Queue()
197	201	# Use subprocess to avoid creation of gRPC threads in main BuildStream process
198	202	# See https://github.com/grpc/grpc/blob/master/doc/fork_support.md for details
199		- process = multiprocessing.Process(target=_queue_wrapper,
200		- args=(_test_push_directory, queue, user_config_file,
201		- project_dir, artifact_dir, artifact_digest))
	203	+ process = multiprocessing.Process(
	204	+ target=_queue_wrapper,
	205	+ args=(_test_push_directory, queue, user_config_file, project_dir,
	206	+ artifact_dir, tmpdir, artifact_digest))
202	207
203	208	try:
204	209	# Keep SIGINT blocked in the child process
...	...	@@ -215,13 +220,17 @@ def test_push_directory(cli, tmpdir, datafiles):
215	220	assert artifact_digest.hash == directory_hash
216	221	assert share.has_object(artifact_digest)
217	222
	223	+ assert os.listdir(os.path.join(str(tmpdir), 'cache', 'tmp')) == []
218	224
219		-def _test_push_directory(user_config_file, project_dir, artifact_dir, artifact_digest, queue):
	225	+
	226	+def _test_push_directory(user_config_file, project_dir, artifact_dir, tmpdir,
	227	+ artifact_digest, queue):
220	228	# Fake minimal context
221	229	context = Context()
222	230	context.load(config=user_config_file)
223	231	context.artifactdir = artifact_dir
224	232	context.set_message_handler(message_handler)
	233	+ context.tmpdir = os.path.join(str(tmpdir), 'cache', 'tmp')
225	234
226	235	# Load the project manually
227	236	project = Project(project_dir, context)
...	...	@@ -273,7 +282,7 @@ def test_push_message(cli, tmpdir, datafiles):
273	282	# See https://github.com/grpc/grpc/blob/master/doc/fork_support.md for details
274	283	process = multiprocessing.Process(target=_queue_wrapper,
275	284	args=(_test_push_message, queue, user_config_file,
276		- project_dir, artifact_dir))
	285	+ project_dir, artifact_dir, tmpdir))
277	286
278	287	try:
279	288	# Keep SIGINT blocked in the child process
...	...	@@ -291,13 +300,16 @@ def test_push_message(cli, tmpdir, datafiles):
291	300	size_bytes=message_size)
292	301	assert share.has_object(message_digest)
293	302
	303	+ assert os.listdir(os.path.join(str(tmpdir), 'cache', 'tmp')) == []
	304	+
294	305
295		-def _test_push_message(user_config_file, project_dir, artifact_dir, queue):
	306	+def _test_push_message(user_config_file, project_dir, artifact_dir, tmpdir, queue):
296	307	# Fake minimal context
297	308	context = Context()
298	309	context.load(config=user_config_file)
299	310	context.artifactdir = artifact_dir
300	311	context.set_message_handler(message_handler)
	312	+ context.tmpdir = os.path.join(str(tmpdir), 'cache', 'tmp')
301	313
302	314	# Load the project manually
303	315	project = Project(project_dir, context)

tests/integration/pullbuildtrees.py

---

...	...	@@ -23,6 +23,7 @@ def default_state(cli, tmpdir, share):
23	23	'artifacts': {'url': share.repo, 'push': False},
24	24	'artifactdir': os.path.join(str(tmpdir), 'artifacts'),
25	25	'cache': {'pull-buildtrees': False},
	26	+ 'tmpdir': os.path.join(str(tmpdir), 'cache', 'tmp'),
26	27	})
27	28
28	29
...	...	@@ -79,6 +80,9 @@ def test_pullbuildtrees(cli, tmpdir, datafiles, integration_cache):
79	80	assert os.path.isdir(buildtreedir)
80	81	default_state(cli, tmpdir, share1)
81	82
	83	+ # Check tmpdir for downloads is cleared
	84	+ assert os.listdir(os.path.join(integration_cache, 'tmp')) == []
	85	+
82	86	# Pull artifact with pullbuildtrees set in user config, then assert
83	87	# that pulling with the same user config doesn't creates a pull job,
84	88	# or when buildtrees cli flag is set.
...	...	@@ -91,6 +95,9 @@ def test_pullbuildtrees(cli, tmpdir, datafiles, integration_cache):
91	95	assert element_name not in result.get_pulled_elements()
92	96	default_state(cli, tmpdir, share1)
93	97
	98	+ # Check tmpdir for downloads is cleared
	99	+ assert os.listdir(os.path.join(integration_cache, 'tmp')) == []
	100	+
94	101	# Pull artifact with default config and buildtrees cli flag set, then assert
95	102	# that pulling with pullbuildtrees set in user config doesn't create a pull
96	103	# job.
...	...	@@ -101,6 +108,9 @@ def test_pullbuildtrees(cli, tmpdir, datafiles, integration_cache):
101	108	assert element_name not in result.get_pulled_elements()
102	109	default_state(cli, tmpdir, share1)
103	110
	111	+ # Check tmpdir for downloads is cleared
	112	+ assert os.listdir(os.path.join(integration_cache, 'tmp')) == []
	113	+
104	114	# Assert that a partial build element (not containing a populated buildtree dir)
105	115	# can't be pushed to an artifact share, then assert that a complete build element
106	116	# can be. This will attempt a partial pull from share1 and then a partial push

tests/sandboxes/storage-tests.py

---

...	...	@@ -3,7 +3,7 @@ import pytest
3	3
4	4	from buildstream._exceptions import ErrorDomain
5	5
6		-from buildstream._artifactcache.cascache import CASCache
	6	+from buildstream._cas import CASCache
7	7	from buildstream.storage._casbaseddirectory import CasBasedDirectory
8	8	from buildstream.storage._filebaseddirectory import FileBasedDirectory
9	9

tests/storage/virtual_directory_import.py

---

...	...	@@ -8,7 +8,7 @@ from tests.testutils import cli
8	8	from buildstream.storage._casbaseddirectory import CasBasedDirectory
9	9	from buildstream.storage._filebaseddirectory import FileBasedDirectory
10	10	from buildstream._artifactcache import ArtifactCache
11		-from buildstream._artifactcache.cascache import CASCache
	11	+from buildstream._cas import CASCache
12	12	from buildstream import utils
13	13
14	14

tests/testutils/artifactshare.py

---

...	...	@@ -11,8 +11,8 @@ from multiprocessing import Process, Queue
11	11	import pytest_cov
12	12
13	13	from buildstream import _yaml
14		-from buildstream._artifactcache.cascache import CASCache
15		-from buildstream._artifactcache.casserver import create_server
	14	+from buildstream._cas import CASCache
	15	+from buildstream._cas.casserver import create_server
16	16	from buildstream._exceptions import CASError
17	17	from buildstream._protos.build.bazel.remote.execution.v2 import remote_execution_pb2
18	18

tests/testutils/runcli.py

---

...	...	@@ -526,7 +526,8 @@ def cli_integration(tmpdir, integration_cache):
526	526	# to avoid downloading the huge base-sdk repeatedly
527	527	fixture.configure({
528	528	'sourcedir': os.path.join(integration_cache, 'sources'),
529		- 'artifactdir': os.path.join(integration_cache, 'artifacts')
	529	+ 'artifactdir': os.path.join(integration_cache, 'artifacts'),
	530	+ 'tmpdir': os.path.join(integration_cache, 'tmp')
530	531	})
531	532
532	533	return fixture
...	...	@@ -573,6 +574,8 @@ def configured(directory, config=None):
573	574	config['builddir'] = os.path.join(directory, 'build')
574	575	if not config.get('artifactdir', False):
575	576	config['artifactdir'] = os.path.join(directory, 'artifacts')
	577	+ if not config.get('tmpdir', False):
	578	+ config['tmpdir'] = os.path.join(directory, 'tmp')
576	579	if not config.get('logdir', False):
577	580	config['logdir'] = os.path.join(directory, 'logs')
578	581

tests/utils/misc.py

---

...	...	@@ -23,7 +23,7 @@ def test_parse_size_over_1024T(cli, tmpdir):
23	23	_yaml.dump({'name': 'main'}, str(project.join("project.conf")))
24	24
25	25	volume_space_patch = mock.patch(
26		- "buildstream._artifactcache.artifactcache.ArtifactCache._get_volume_space_info_for",
	26	+ "buildstream._artifactcache.ArtifactCache._get_volume_space_info_for",
27	27	autospec=True,
28	28	return_value=(1025 * TiB, 1025 * TiB)
29	29	)

—
View it on GitLab.
You're receiving this email because of your account on gitlab.com. If you'd like to receive fewer emails, you can adjust your notification settings.