update: in nodriver mode, avoid loading proc, users and interfaces related informations

[FedeDP]

Signed-off-by: Federico Di Pierro [email protected]

What type of PR is this?

Uncomment one (or more) /kind <> lines:

/kind cleanup

Any specific area of the project related to this PR?

/area libscap

/area libsinsp

What this PR does / why we need it:

In nodriver mode, skip loading proc, users and interfaces related informations as all event sources will be system-external.

Which issue(s) this PR fixes:

Possibly: falcosecurity/falco#1757

Fixes #

Special notes for your reviewer:

Does this PR introduce a user-facing change?:

NONE

[poiana]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by: FedeDP
To complete the pull request process, please assign gnosek after the PR has been reviewed.
You can assign the PR to them by writing /assign @gnosek in a comment when ready.

The full list of commands accepted by this bot can be found here.

Needs approval from an approver in each of these files:

• OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[poiana]

Hi @FedeDP. Thanks for your PR.

I'm waiting for a falcosecurity member to verify that this patch is reasonable to test. If it is, they should reply with /ok-to-test on its own line. Until that is done, I will not automatically test new commits in this PR, but the usual testing commands by org members will still work. Regular contributors should join the org to skip this step.

Once the patch is verified, the new status will be reflected by the ok-to-test label.

I understand the commands that are listed here.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.

[leogr]

/ok-to-test

[mstemm]

Don't you still want the initial process/user/interface lists in nodriver mode? I'm not sure I understand why we would not want them.

[FedeDP]

Don't you still want the initial process/user/interface lists in nodriver mode? I'm not sure I understand why we would not want them.

What's the use for that? I mean, is there any particular use case where we want that in nodriver mode?

EDIT: surely i can be missing a piece there :)

[mstemm]

I thought nodriver mode still attempted to get a reconstructed view of the world and a minimal set of events that could be written to a capture file. (I don't remember the details). Do you know?

[FedeDP]

I thought nodriver mode still attempted to get a reconstructed view of the world and a minimal set of events that could be written to a capture file. (I don't remember the details). Do you know?

As nodriver mode has no events from syscalls. there is no real "world's view" in that sense.
We always dump:

• scap_write_machine_info
• scap_write_iflist
• scap_write_userlist
• scap_write_proclist
• scap_write_fdlist

when opening the scap file. With the implementation, all of these will be empty (except for machine_info possibly).
Note, moreover, that these scap_write_ functions already have a

//
// No machine info on disk if the source is a plugin
//
if(handle->m_mode == SCAP_MODE_PLUGIN)
{
	return SCAP_SUCCESS;
}

IE: we are already avoiding dumping initial world for plugins; this is what we need to do not only for plugins, but whenever internal event source (ie: syscalls) is disabled.
I may rearrange the code with a more explicit check (eg: adding an helper method on scap is_system_event_source_enabled() or something similar).

[mstemm]

One big difference between nodriver mode and plugin mode is that when in plugin mode, the only events and state are within the plugin. All of libsinsp is effectively disabled.

That's not true for nodriver mode, where the goal is to still obtain process and thread level state of the system.

Let me get more familiar with nodriver mode and then comment more.

[mstemm]

I double-checked and nodriver mode depends on reading this info from /proc, so I think if the goal is to come up with a "light" mode that allowed only k8s audit logs and no syscalls, we should come up with another mode/solution, perhaps at the falco level instead of the libs level, to fix this.

[FedeDP]

I double-checked and nodriver mode depends on reading this info from /proc, so I think if the goal is to come up with a "light" mode that allowed only k8s audit logs and no syscalls, we should come up with another mode/solution, perhaps at the falco level instead of the libs level, to fix this.

I think that when (if?) k8s audit logs will become a plugin, the issue will be fixed in any case; it's ok for me to close this PR and wait for a proper plugin implementation.

[FedeDP]

Closing because nodriver mode still needs a system view.
Issue itself (falcosecurity/falco#1757) will be fixed once k8s audit log becomes a plugin.