agent/api/task/task.go (12 lines): - line 206: // TODO DesiredStatusUnsafe should probably be private with appropriately written - line 217: // TODO KnownStatusUnsafe should probably be private with appropriately written - line 236: // TODO SentStatusUnsafe should probably be private with appropriately written - line 368: // TODO: Have EBS volumes use the DockerVolumeConfig to create the mountpoint - line 555: // TODO [SC]: ServiceConnectConfig will come from ACS. Adding this here for dev/testing purposes only Remove when - line 563: // TODO [SC]: initDummyServiceConnectConfig is for dev testing only, remove it when final SC model from ACS is in place - line 572: // TODO [SC]: This is for dev testing only, remove it when final SC model from ACS is in place - line 1791: // TODO [SC] - Move this as well as 'dockerExposedPorts' SC-specific logic into a separate file - line 2710: // TODO: Create a mapping of resource status to the corresponding task status and use it here - line 2855: // TODO: what's the point of locking if we are returning a pointer? - line 3509: // TODO: as more daemons come online, we'll want a generic handler these bool checks and payload handling - line 3794: // TODO: Somehow track this on a task level (i.e. obtain the managed daemon image name from task arn and then find the corresponding container with the image name) agent/api/container/container.go (5 lines): - line 133: // TODO: [ecs-exec] Change variable name from Status to KnownStatus in future PR to avoid noise - line 240: // TODO DesiredStatusUnsafe should probably be private with appropriately written - line 248: // TODO KnownStatusUnsafe should probably be private with appropriately written - line 285: // TODO SentStatusUnsafe should probably be private with appropriately written - line 1266: // TODO return error to differentiate between error and default mode . agent/dockerclient/dockerapi/docker_client.go (5 lines): - line 209: // TODO ListPluginsWithFilters can be removed since ListPlugins takes in filters - line 258: // TODO Remove clientfactory field once all API calls are migrated to sdkclientFactory - line 308: // TODO Remove clientfactory parameter once migration to Docker SDK is complete. - line 723: // TODO Remove ContainerInspect call - line 1036: // TODO: move the error check into StartListening() to keep event streaming and error handling in one place. agent/ebs/watcher.go (4 lines): - line 44: // TODO: The dataClient will be used to save to agent's data client as well as start the ACK timer. This will be added once the data client functionality have been added - line 50: // TODO: The dockerTaskEngine.stateChangeEvent will be used to send over the state change event for EBS attachments once it's been found and mounted/resize/format. - line 301: // TODO: Add the EBS volume to data client - line 319: // TODO: Remove the EBS volume from the data client. agent/stats/engine.go (3 lines): - line 242: // TODO ensure that this is done only once per engine object - line 1058: // TODO [SC]: Check if task is service-connect enabled - line 1104: // TODO: Remove the CSI client from the stats engine and just always have the CSI client created agent/engine/execcmd/manager_init_task.go (3 lines): - line 68: // TODO: [ecs-exec] Should we validate the ssm agent binaries & certs are valid and fail here if they're not? (bind mount will succeed even if files don't exist in host) - line 118: // TODO: [ecs-exec] This requirement will be removed for SSM agent V2 - line 155: // TODO [ecs-exec] : verify that returning the default session limit (2) is ok in case of any errors, misconfiguration ecs-agent/daemonimages/csidriver/mounter/safe_mounter_windows.go (3 lines): - line 82: // TODO: Confirm that fstype is empty. - line 377: // TODO: consider partitions and choose the right partition. - line 394: // TODO: Accept the filesystem and other options agent/engine/task_manager.go (3 lines): - line 257: // TODO [SC]: We need to also tear down pause containets in bridge mode for SC-enabled tasks - line 261: // TODO: make this idempotent on agent restart - line 1481: // TODO we should probably panic here agent/engine/execcmd/manager_start.go (2 lines): - line 67: //TODO: [ecs-exec] retry only for certain exit codes? - line 222: if !inspect.Running { //TODO: [ecs-exec] retry only for certain exit codes? ecs-agent/manageddaemon/managed_daemon_linux.go (2 lines): - line 40: // TODO parse taskdef json files in parameterized dir ie /deps/daemons - line 41: // TODO validate that each daemon's layers are loaded or that daemon has a corresponding image tar agent/engine/docker_task_engine.go (2 lines): - line 514: // TODO: [ecs-exec]add jitter between containers to not overload docker with top calls - line 2252: // TODO: Add a state to the apicontainer.Container for the status of the metadata file (Whether it needs update) and ecs-agent/tcs/client/client.go (2 lines): - line 269: // TODO [SC]: Load test and profile this since BuildJSON results in lot of CPU and memory consumption. - line 504: // TODO: Modify this to use SignHTTPRequest() once TCS has been migrated to use AWS SDK Go V2 agent/engine/docker_image_manager.go (2 lines): - line 140: // TODO add setter and getter for modifying this - line 310: // TODO: change image createdTime to image lastUsedTime when docker support it in the future agent/app/agent.go (2 lines): - line 529: // TODO add EBS watcher to async routines - line 684: // TODO Handle default cluster in a sane and unified way across the codebase ecs-agent/netlib/model/networkinterface/networkinterface.go (2 lines): - line 420: // TODO: Primary field is not available yet so set the first one to be primary for now. - line 460: // TODO: check subnet gateway for IPv6 ecs-agent/wsclient/client.go (2 lines): - line 179: // TODO: Remove this once both ACS and TCS implementations have migrated over to AWS SDK Go V2. - line 256: //TODO: In order to get rid of the check - agent/handlers/v2/response.go (2 lines): - line 134: // TODO: remove includeV4Metadata from NewContainerResponseFromState/NewContainerResponse - line 156: // TODO: remove includeV4Metadata from NewContainerResponse agent/api/statechange.go (2 lines): - line 518: // TODO: This check already exists in NewContainerStateChangeEvent and shouldn't be repeated here; remove after verifying - line 530: // TODO: This check is probably redundant as String() method never returns "DEAD"; remove after verifying agent/doctor/docker_runtime_healthcheck.go (1 line): - line 59: // TODO pass in context as an argument agent/acs/session/payload_responder.go (1 line): - line 344: // TODO: This will only work if there's one EBS volume per task. If we there is a case where we have multi-attach for a task, this needs to be modified ecs-agent/tcs/client/error.go (1 line): - line 44: // TODO: Populate this list when the json is updated with actual exceptions. misc/volumes-test/build.ps1 (1 line): - line 15: # TODO be smart about this and only rebuild if necessary agent/engine/data.go (1 line): - line 59: // TODO: Will need to clean up all of the STOPPED managed daemon tasks agent/config/config_unix.go (1 line): - line 174: // TODO feat:IPv6-only - Remove lint rule below agent/taskresource/credentialspec/credentialspec_windows.go (1 line): - line 468: // TODO: Use registry to setup credentialspec resource location agent/utils/ephemeral_ports.go (1 line): - line 78: // TODO: implement a port manager that does synchronization and integrates with a configurable option to modify ephemeral range agent/engine/common_testutil.go (1 line): - line 102: // TODO: Move integ tests away from relying on the statechange channel for agent/eventhandler/task_handler.go (1 line): - line 330: // TODO: wire in the context here. Else, we have go routine leaks in tests ecs-agent/utils/sign_http_request.go (1 line): - line 77: // TODO: Remove this once both ACS and TCS have been migrated over to AWS SDK Go V2 ecs-init/docker/docker.go (1 line): - line 235: // TODO pagination agent/api/testutils/container_equal.go (1 line): - line 28: // TODO: use reflection along with `equal:"unordered"` annotations on slices to ecs-init/volumes/state_manager.go (1 line): - line 31: // TODO: get this value from an env var agent/handlers/task_server_setup.go (1 line): - line 197: // TODO: Will need to be called in taskServerSetup once all of the handlers have been implemented ecs-agent/logger/eventlog_windows.go (1 line): - line 26: TODO: Make this whole thing better agent/statemanager/state_manager_windows.go (1 line): - line 146: // TODO Investigate performance differences between FlushFileBuffers and unbuffered IO. agent/data/models/task_models.go (1 line): - line 74: // TODO: update this model when introducing first actual transformation function agent/stats/service_connect_linux.go (1 line): - line 56: // TODO [SC]: Add retries on failure to retrieve service connect stats agent/config/config.go (1 line): - line 242: // TODO feat:IPv6-only - Enable when launching IPv6-only support packaging/suse/amazon-ecs-init.spec (1 line): - line 17: # TODO: replace with generic rpm. Need to verify the latter works on EC2. agent/engine/daemonmanager/daemon_manager.go (1 line): - line 149: // TODO update User in raw config to use either runAs user or runAsRoot from managed daemon config agent/dockerclient/dockerapi/types.go (1 line): - line 107: // TODO Remove type when migration is complete agent/engine/serviceconnect/manager_linux.go (1 line): - line 318: // TODO [SC] - Move the function here misc/windows-deploy/amazon-ecs-agent.ps1 (1 line): - line 20: ## TODO: This is a very basic logging function. We should make it better, or replace it with agent/app/agent_windows.go (1 line): - line 181: // TODO: Pre-start hooks go here (unclear if we need any yet) agent/data/transformationfunctions/tasktf.go (1 line): - line 35: // TODO: edit this function when introducing first actual transformation function, and add unit test agent/eni/networkutils/utils_linux.go (1 line): - line 93: // a timeout here. TODO: Fix RetryWithBackoffCtx to return ctx.Err() on context Done() ecs-agent/wsclient/mock/utils/utils.go (1 line): - line 39: // TODO replace with gomock ecs-agent/credentials/providers/rotating_shared_credentials_provider_v2.go (1 line): - line 31: // TODO (@tiffwang): Remove V2 suffix after the credentials package is ecs-agent/utils/httpproxy/proxy.go (1 line): - line 297: // TODO: Consider removing this check after verifying performance is okay. agent/dockerclient/dockerapi/docker_events_buffer.go (1 line): - line 24: // TODO add support for filter in go-dockerclient ecs-init/volumes/ecs_volume_driver.go (1 line): - line 27: // TODO: might be a good idea to check the mount point with a tool like mountpoint instead of matching error message. agent/engine/execcmd/manager_init_task_linux.go (1 line): - line 78: // TODO: [ecs-exec] seelog config needs to be implemented following a similar approach to ss, config agent/taskresource/asmauth/asmauth.go (1 line): - line 63: // TODO: Refactor this struct so that each ASMAuthData gets associated with agent/handlers/v4/response.go (1 line): - line 165: // TODO this is hard-coded to `0` for now. Once backend starts populating agent/app/agent_capability.go (1 line): - line 244: // TODO: gate this on docker api version when ecs supported docker includes agent/utils/ephemeral_ports_windows.go (1 line): - line 15: // TODO: instead of sticking to defaults, run netsh commands on the host to get the ranges. agent/eni/watcher/watcher.go (1 line): - line 204: // a timeout here. TODO: Fix RetryWithBackoffCtx to return ctx.Err() on context Done()