Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Add missing project SSH key to on-prem instances #1716

Merged
merged 2 commits into from
Sep 24, 2024
Merged

Add missing project SSH key to on-prem instances #1716

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
86764e6
Add missing project SSH key to on-prem instances
un-def Sep 23, 2024
2150dc3
Add TODO
un-def Sep 24, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
21 changes: 11 additions & 10 deletions src/dstack/_internal/server/background/tasks/process_instances.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -230,9 +230,12 @@ async def _add_remote(instance: InstanceModel) -> None:
)
return

authorized_keys = [pk.public.strip() for pk in remote_details.ssh_keys]
authorized_keys.append(instance.project.ssh_public_key.strip())

try:
future = asyncio.get_running_loop().run_in_executor(
None, _deploy_instance, remote_details, pkeys
None, _deploy_instance, remote_details, pkeys, authorized_keys
)
deploy_timeout = 20 * 60 # 20 minutes
result = await asyncio.wait_for(future, timeout=deploy_timeout)
Expand Down Expand Up @@ -317,7 +320,9 @@ async def _add_remote(instance: InstanceModel) -> None:


def _deploy_instance(
remote_details: RemoteConnectionInfo, pkeys: List[PKey]
remote_details: RemoteConnectionInfo,
pkeys: List[PKey],
authorized_keys: List[str],
) -> Tuple[HealthStatus, Dict[str, Any]]:
with get_paramiko_connection(
remote_details.ssh_user, remote_details.host, remote_details.port, pkeys
Expand All @@ -328,17 +333,11 @@ def _deploy_instance(

# Execute pre start commands
shim_pre_start_commands = get_shim_pre_start_commands(runner_build)
run_pre_start_commands(
client,
shim_pre_start_commands,
authorized_keys=[pk.public.strip() for pk in remote_details.ssh_keys],
)
run_pre_start_commands(client, shim_pre_start_commands, authorized_keys)
logger.debug("The script for installing dstack has been executed")

# Upload envs
shim_envs = get_shim_env(
runner_build, authorized_keys=[sk.public for sk in remote_details.ssh_keys]
)
shim_envs = get_shim_env(runner_build, authorized_keys)
try:
fleet_configuration_envs = remote_details.env.as_dict()
except ValueError as e:
Expand Down Expand Up @@ -538,6 +537,8 @@ async def _check_instance(instance: InstanceModel) -> None:
return

ssh_private_key = instance.project.ssh_private_key
# TODO: Drop this logic and always use project key once it's safe to assume that most on-prem
# fleets are (re)created after this change: https://github.com/dstackai/dstack/pull/1716
if instance.remote_connection_info is not None:
remote_conn_info: RemoteConnectionInfo = RemoteConnectionInfo.__response__.parse_raw(
instance.remote_connection_info
Expand Down
2 changes: 2 additions & 0 deletions src/dstack/_internal/server/background/tasks/process_running_jobs.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -134,6 +134,8 @@ async def _process_running_job(session: AsyncSession, job_model: JobModel):
)

server_ssh_private_key = project.ssh_private_key
# TODO: Drop this logic and always use project key once it's safe to assume that most on-prem
# fleets are (re)created after this change: https://github.com/dstackai/dstack/pull/1716
if (
job_model.instance is not None
and job_model.instance.remote_connection_info is not None
Expand Down
5 changes: 5 additions & 0 deletions src/dstack/_internal/server/services/jobs/__init__.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -149,6 +149,8 @@ async def stop_runner(session: AsyncSession, job_model: JobModel):
)
instance: Optional[InstanceModel] = res.scalar()

# TODO: Drop this logic and always use project key once it's safe to assume that most on-prem
# fleets are (re)created after this change: https://github.com/dstackai/dstack/pull/1716
if instance and instance.remote_connection_info is not None:
remote_conn_info: RemoteConnectionInfo = RemoteConnectionInfo.__response__.parse_raw(
instance.remote_connection_info
Expand Down Expand Up @@ -210,6 +212,9 @@ async def process_terminating_job(session: AsyncSession, job_model: JobModel):
jpd = JobProvisioningData.__response__.parse_raw(job_model.job_provisioning_data)
logger.debug("%s: stopping container", fmt(job_model))
ssh_private_key = instance.project.ssh_private_key
# TODO: Drop this logic and always use project key once it's safe to assume that
# most on-prem fleets are (re)created after this change:
# https://github.com/dstackai/dstack/pull/1716
if instance and instance.remote_connection_info is not None:
remote_conn_info: RemoteConnectionInfo = (
RemoteConnectionInfo.__response__.parse_raw(
Expand Down