Code
Hub
Workspaces
Following
Trending
Connect
MCP
copy
Create free account
hub
/
github.com/LMCache/LMCache
/ functions
Functions
13,371 in github.com/LMCache/LMCache
⨍
Functions
13,371
◇
Types & classes
2,044
↳
Endpoints
181
↓ 12 callers
Function
add_handler_helper
Register a handler with the message queue server. Args: server: The message queue server. request_type: The request type to handl
lmcache/v1/multiprocess/server.py:142
↓ 12 callers
Function
apply_remote_configs
Apply remote configuration to LMCacheEngineConfig. This function extracts the 'configs' field from the remote response and applies each confi
lmcache/v1/config_base.py:798
↓ 12 callers
Method
batched_contains
Check how many prefix keys exist via single batch_exists RPC. Args: keys: Keys to check in prefix order. pin: If True
lmcache/v1/storage_backend/maru_backend.py:627
↓ 12 callers
Method
close
(self)
tests/v1/shm_allocator/verify_shmfile_tcp.py:103
↓ 12 callers
Method
create_metrics
Create a :class:`Metrics` with default handlers pre-registered. Handlers are configured from ``args.format``, ``args.output``, and ``
lmcache/cli/commands/base.py:93
↓ 12 callers
Function
create_test_config
Create a test configuration for connector plugin testing.
tests/v1/storage_backend/test_remote_storage_plugin.py:116
↓ 12 callers
Method
debug_get_stored_object_count
Return the number of objects currently stored in the mock adapter. This method is intended for testing and debugging only.
lmcache/v1/distributed/l2_adapters/mock_l2_adapter.py:292
↓ 12 callers
Method
delete
Delete the given keys from L1 cache. Args: keys: The list of object keys to delete. Returns: A dictionary ma
lmcache/v1/distributed/l1_manager.py:662
↓ 12 callers
Function
derefFloat64
derefFloat64 returns the value pointed to by p, or def if p is nil.
operator/internal/resources/helpers.go:44
↓ 12 callers
Function
detect_format
Recover a contiguous view, then discover the format + canonical kv_caches. Returns ``(engine_kv_format, normalized_kv_caches)``. Callers must use
lmcache/v1/gpu_connector/kv_format/detection.py:22
↓ 12 callers
Method
exists
(self, key: CacheEngineKey)
tests/v1/shm_allocator/shmfile_connector.py:207
↓ 12 callers
Method
exists_many
Check whether keys exist, optionally acquiring external locks. Args: keys: Keys to look up. lock: If ``True``, increm
lmcache/v1/storage_backend/dax/core.py:283
↓ 12 callers
Function
findContainer
findContainer returns the named container from the pod, or nil.
operator/internal/webhook/pod_injector_test.go:170
↓ 12 callers
Function
get_dtype
Return the dtype for a layer (defaults to layer 0).
lmcache/v1/gpu_connector/utils.py:314
↓ 12 callers
Method
get_dtype
Get the dtype of the MemoryObj.
lmcache/v1/memory_management.py:237
↓ 12 callers
Method
get_evict_candidates
( self, cache_dict: dict[KeyType, Any], num_candidates: int = 1, )
lmcache/v1/storage_backend/cache_policy/lfu.py:80
↓ 12 callers
Method
get_eviction_candidates
Get a list of eviction candidates without creating eviction actions. Useful for querying what would be evicted. Note:
lmcache/v1/distributed/eviction_policy/lru.py:220
↓ 12 callers
Method
get_kv_count
Get total count of KV chunks.
lmcache/v1/cache_controller/utils.py:182
↓ 12 callers
Method
get_shape
Get the shape of the MemoryObj.
lmcache/v1/memory_management.py:231
↓ 12 callers
Method
get_temp_kernel_group_buffer
( self, batch_idx: int, kernel_group_idx: int )
tests/v1/platform/test_cache_context_dispatch.py:98
↓ 12 callers
Method
get_worker_kv_keys
Get all KV keys for a specific worker and location.
lmcache/v1/cache_controller/utils.py:636
↓ 12 callers
Method
help
(cls)
tests/v1/distributed/test_l2_adapter_factory.py:310
↓ 12 callers
Function
make_descriptor
Create an AdapterDescriptor for testing.
tests/v1/distributed/test_store_controller.py:159
↓ 12 callers
Function
make_layout
Create a small MemoryLayoutDesc for testing.
tests/v1/distributed/test_store_controller.py:62
↓ 12 callers
Function
make_mock_config
()
tests/v1/distributed/test_runtime_adapters.py:79
↓ 12 callers
Function
parse_mixed_slot_mapping
Parse mixed format slot_mapping string. Supports two formats: 1. Single numbers: "1,2,3,17,19" 2. Range format: "[9,12]" (represents 9,10
lmcache/utils.py:187
↓ 12 callers
Function
plan_to_indices
Convert a Bitmap-based load plan to index lists for easy assertion.
tests/v1/distributed/test_prefetch_policy.py:49
↓ 12 callers
Method
post_init
Post-initialization after KV caches are registered.
lmcache/v1/manager.py:181
↓ 12 callers
Function
publish_call_event
Publish one ``TRACE_CALL`` event. Used by :func:`enable_tracing` and by manual instrumentation points (e.g. context-manager enter/exit) that
lmcache/v1/mp_observability/trace/decorator.py:59
↓ 12 callers
Method
query
Check if the future is done. Returns: bool: True if the future is done, False otherwise.
lmcache/v1/multiprocess/futures.py:17
↓ 12 callers
Method
read
(self, *_, **__)
tests/v1/storage_backend/test_gds_backend.py:425
↓ 12 callers
Function
snapshot
Capture counters before the test; yield a callable that returns deltas.
tests/v1/mp_observability/subscribers/metrics/test_l2.py:110
↓ 12 callers
Method
split_layers
Split the key into multiple keys for each layer
lmcache/utils.py:458
↓ 12 callers
Method
to_encoded_object_key
Return the JSON-safe :class:`EncodedObjectKey` projection.
lmcache/v1/distributed/api.py:100
↓ 12 callers
Function
wait_for_event_fd
(event_fd: int, timeout: float = 5.0)
tests/v1/distributed/test_dax_l2_adapter.py:98
↓ 11 callers
Function
BuildDaemonSet
BuildDaemonSet constructs a DaemonSet for the given LMCacheEngine.
operator/internal/resources/daemonset.go:48
↓ 11 callers
Function
CreateGPUConnector
Create a GPU Connector based on the configuration and metadata. Args: config: The LMCache engine configuration. metadata: Th
lmcache/v1/gpu_connector/__init__.py:60
↓ 11 callers
Function
_adapter_status
( backend: str, adapter_index: int, status: Optional[dict[str, object]] = None, supported_oper
tests/v1/multiprocess/http_apis/test_reconfigure_api.py:20
↓ 11 callers
Function
_build_manager
Build a real :class:`KVLayerGroupsManager` from synthetic tensors.
tests/v1/platform/test_gpu_cache_context.py:90
↓ 11 callers
Function
_client
(sm: _FakeStorageManager)
tests/v1/multiprocess/http_apis/test_reconfigure_api.py:74
↓ 11 callers
Function
_create_test_memory_obj
Allocate a test memory object with deterministic random data. Args: local_backend: The local CPU backend for allocation. seed: Ra
tests/v1/storage_backend/test_valkey_connector.py:125
↓ 11 callers
Function
_events_body
(events: list[dict], instance_id: str = "test-server")
tests/v1/mp_coordinator/test_l2_api.py:41
↓ 11 callers
Function
_grp
Cast a tuple of test fakes to the production MemoryObjGroup type.
tests/v1/distributed/serde/test_asym_k16_v8_v_only.py:67
↓ 11 callers
Function
_instance
Build a pure-membership MPInstance for tests.
tests/v1/mp_coordinator/test_registry.py:11
↓ 11 callers
Function
_make_config
( device_path: str, *, slot_bytes: int = 64 * 1024, capacity_bytes: int = 0, io_engine: st
tests/v1/distributed/test_raw_block_l2_adapter.py:134
↓ 11 callers
Function
_make_event
(event_type: EventType, keys: list)
tests/v1/mp_observability/subscribers/metrics/test_l1.py:28
↓ 11 callers
Function
_make_keys
Create *n* distinct CacheEngineKeys (chunk_hash = 0 … n-1).
tests/v1/test_nixl_batched_contains.py:43
↓ 11 callers
Function
_make_keys
Create a list of placeholder key objects with a cache_salt attribute.
tests/v1/mp_observability/subscribers/metrics/test_l1.py:23
↓ 11 callers
Function
_make_metadata
()
tests/v1/storage_backend/test_nixl_shared_pool.py:107
↓ 11 callers
Function
_make_mock_memory_obj
Create a mock MemoryObj that tracks ref_count_down calls.
tests/v1/test_cache_engine.py:1671
↓ 11 callers
Method
_on_retrieve_end
(self, event: Event)
lmcache/v1/mp_observability/subscribers/metrics/engine.py:52
↓ 11 callers
Method
_on_store_completed
(self, event: Event)
lmcache/v1/mp_observability/subscribers/metrics/l2.py:147
↓ 11 callers
Method
_on_store_submitted
(self, event: Event)
lmcache/v1/mp_observability/subscribers/metrics/l2.py:140
↓ 11 callers
Function
_run_store
(adapter: RawBlockL2Adapter, keys, objects)
tests/v1/distributed/test_raw_block_l2_adapter.py:207
↓ 11 callers
Function
_store_submitted
( task_id: int, t: float, adapter_index: int = 0, l2_name: str = "fs", total_bytes: int =
tests/v1/mp_observability/subscribers/metrics/test_l2_throughput.py:34
↓ 11 callers
Function
_t
(*shape: int)
tests/v1/gpu_connector/test_kv_format_detection.py:25
↓ 11 callers
Function
_t
(*shape: int)
tests/v1/gpu_connector/test_kv_format_specs.py:29
↓ 11 callers
Function
_wait_event_fd
Wait for an eventfd to be signaled.
examples/lmc_external_l2_adapter/tests/test_plugin.py:137
↓ 11 callers
Method
add_l2_adapter
Blocking function to add a new L2 adapter at runtime. Thread-safe. Args: config: The adapter configuration. Returns:
lmcache/v1/distributed/storage_manager.py:817
↓ 11 callers
Function
argsHasFlagValue
--- test arg helpers (mirror the package's two-token / =-token recognition) --- argsHasFlagValue reports whether args carries flag with the given valu
operator/internal/webhook/pod_injector_test.go:545
↓ 11 callers
Method
available_for_read
Check if the object is available for read. Returns: True if the object is not write-locked, False otherwise.
lmcache/v1/distributed/l1_manager.py:50
↓ 11 callers
Method
available_for_write
Check if the object is available for write. Returns: True if the object is not write-locked and has no read locks and
lmcache/v1/distributed/l1_manager.py:58
↓ 11 callers
Function
check_paged_kv_cache_equal_with_mla
check whether two paged kv caches are the same at slot_mapping when use mla
tests/v1/utils.py:607
↓ 11 callers
Method
close
(self)
lmcache/v1/multiprocess/mq.py:868
↓ 11 callers
Function
create_cache_key
Create a cache key for testing.
tests/v1/multiprocess/test_cache_server.py:132
↓ 11 callers
Function
create_l2_adapter
Create an L2 adapter from its config via the factory registry. Args: config: The adapter-specific config object. l1_memory_de
lmcache/v1/distributed/l2_adapters/__init__.py:37
↓ 11 callers
Function
create_memory_obj
(size: int = 16, fill_value: float = 1.0)
tests/v1/distributed/test_hfbucket_l2_adapter.py:144
↓ 11 callers
Function
create_test_config
Create a test configuration for LocalCPUBackend.
tests/v1/storage_backend/test_local_cpu_backend.py:44
↓ 11 callers
Function
create_test_config
Create a test configuration for LocalDiskBackend.
tests/v1/storage_backend/test_local_disk_backend.py:43
↓ 11 callers
Method
create_tracker
Create FullSyncTracker with registry.
tests/v1/cache_controller/conftest.py:147
↓ 11 callers
Function
escapeHtmlStr
(str)
lmcache/lmcache_frontend/static/js/app.js:706
↓ 11 callers
Method
get
Retrieve the full chunk associated with ``key``.
lmcache/v1/storage_backend/connector/hfbucket_connector.py:265
↓ 11 callers
Method
get
(self, key: CacheEngineKey)
lmcache/v1/storage_backend/connector/eic_connector.py:425
↓ 11 callers
Function
get_block_size
Return the block size (tokens per block) for layer ``layer_idx``. ``layer_idx`` is honoured only for per-layer formats where BS may differ ac
lmcache/v1/gpu_connector/utils.py:244
↓ 11 callers
Method
get_blocking
(self, key: CacheEngineKey)
lmcache/v1/storage_backend/plugins/rust_raw_block_backend.py:581
↓ 11 callers
Function
get_cache_policy
Factory function to get the cache policy instance based on the policy name. Args: policy_name: Name of the cache policy (case-insens
lmcache/v1/storage_backend/cache_policy/__init__.py:21
↓ 11 callers
Method
get_instance
Get or create the singleton, incrementing the ref count. Returns: ClientPollingLoop: The shared polling loop instance.
lmcache/v1/multiprocess/mq.py:152
↓ 11 callers
Function
get_num_layers
Return the number of layers from ``kv_caches``.
lmcache/v1/gpu_connector/utils.py:226
↓ 11 callers
Method
get_row_key
(self, key: CacheEngineKey)
lmcache/v1/storage_backend/connector/bigtable_schema.py:16
↓ 11 callers
Method
handle_orchestration_message
(self, msg: OrchMsg)
lmcache/v1/cache_controller/controller_manager.py:210
↓ 11 callers
Method
hotplug_status
Return runtime status for this DAX adapter. Returns: JSON-serializable status for the stable adapter facade and every
lmcache/v1/distributed/l2_adapters/dax_l2_adapter.py:664
↓ 11 callers
Method
init_mutable_mapping
(self)
lmcache/v1/storage_backend/cache_policy/lfu.py:35
↓ 11 callers
Method
insert
Insert a new entry, or refresh if already present. O(1).
lmcache/tools/cache_simulator/lru_cache.py:49
↓ 11 callers
Method
keys_for
(self, token_ids)
tests/v1/test_hidden_state_store.py:82
↓ 11 callers
Method
mark_kv_present
(self, keys)
tests/v1/test_hidden_state_store.py:85
↓ 11 callers
Method
memcheck
(self)
lmcache/v1/memory_management.py:2355
↓ 11 callers
Method
on_request_finished
(self, request_id: str, output: str)
tests/cli/commands/bench/engine_bench/workloads/test_base_workload.py:39
↓ 11 callers
Method
pin
Pin a key in storage.
tests/v1/storage_backend/test_storage_plugin.py:98
↓ 11 callers
Method
record
Record one replayed call. Args: qualname: Qualified name of the replayed function. latency_s: Elapsed seconds for the
lmcache/cli/commands/trace/_stats.py:134
↓ 11 callers
Method
register
Register a handler for *qualname*. Args: qualname: Fully-qualified call-site name exactly matching the ``qualname
lmcache/cli/commands/trace/_dispatch.py:82
↓ 11 callers
Method
register_fd
Mock register_fd - no-op for testing.
tests/conftest.py:552
↓ 11 callers
Method
report_status
Return a status dict aggregating all sub-component statuses.
lmcache/v1/distributed/storage_manager.py:931
↓ 11 callers
Function
run
()
tests/v1/storage_backend/test_pd_backend_async.py:543
↓ 11 callers
Method
select_load_plan
Assign each key to the first adapter (by index) that has it. Args: keys: Full list of keys being prefetched.
lmcache/v1/distributed/storage_controllers/prefetch_policy.py:135
↓ 11 callers
Function
shm_create_readwrite
Create a new shared-memory segment and return its mmap address. Mirrors the previous ``shm_open(O_CREAT|O_EXCL) + ftruncate + mmap`` sequence
lmcache/v1/multiprocess/posix_shm.py:122
↓ 11 callers
Function
shm_munmap
Best-effort release of a previously mapped segment by address. The underlying mmap is closed exactly once; subsequent calls with the same add
lmcache/v1/multiprocess/posix_shm.py:171
↓ 11 callers
Method
start
(self)
lmcache/v1/multiprocess/mq.py:857
↓ 11 callers
Method
unregister
Unregister a client's DEALER socket from the shared poller. Blocks until the loop thread has completed the unregistration. Args:
lmcache/v1/multiprocess/mq.py:195
↓ 11 callers
Function
validate_group_size
Validation helper used by callers and implementations alike. Pure arity check: the element type is irrelevant, so ``group`` is typed ``Sequen
lmcache/v1/distributed/serde/multi.py:273
↓ 11 callers
Method
validate_nixl_backend
(backend: str, device: str)
lmcache/v1/storage_backend/nixl_storage_backend.py:122
← previous
next →
501–600 of 13,371, ranked by callers