Skip to content

Make events with missing parents go through eventcache #318

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.

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

Merged
merged 5 commits into from
Aug 12, 2022
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
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
16 changes: 16 additions & 0 deletions api/v1/tetragon/tetragon_ext.go
Original file line number Diff line number Diff line change
Expand Up @@ -47,3 +47,19 @@ func (x *ProcessKprobe) SetProcess(p *Process) {
func (x *ProcessTracepoint) SetProcess(p *Process) {
x.Process = p
}

func (x *ProcessExec) SetParent(p *Process) {
x.Parent = p
}

func (x *ProcessExit) SetParent(p *Process) {
x.Parent = p
}

func (x *ProcessKprobe) SetParent(p *Process) {
x.Parent = p
}

func (x *ProcessTracepoint) SetParent(p *Process) {
x.Parent = p
}
1 change: 1 addition & 0 deletions pkg/eventcache/eventcache.go
Original file line number Diff line number Diff line change
Expand Up @@ -50,6 +50,7 @@ type Cache struct {
var (
ErrFailedToGetPodInfo = errors.New("failed to get pod info")
ErrFailedToGetProcessInfo = errors.New("failed to get process info")
ErrFailedToGetParentInfo = errors.New("failed to get parent info")
)

// Generic internal lookup happens when events are received out of order and
Expand Down
66 changes: 51 additions & 15 deletions pkg/grpc/exec/exec.go
Original file line number Diff line number Diff line change
Expand Up @@ -14,7 +14,6 @@ import (
"github.com/cilium/tetragon/pkg/ktime"
"github.com/cilium/tetragon/pkg/logger"
"github.com/cilium/tetragon/pkg/metrics/errormetrics"
"github.com/cilium/tetragon/pkg/metrics/processexecmetrics"
"github.com/cilium/tetragon/pkg/option"
"github.com/cilium/tetragon/pkg/process"
readerexec "github.com/cilium/tetragon/pkg/reader/exec"
Expand All @@ -37,11 +36,7 @@ func GetProcessExec(proc *process.ProcessInternal) *tetragon.ProcessExec {
processId := tetragonProcess.ExecId

parent, err := process.Get(parentId)
if err != nil {
errormetrics.ErrorTotalInc(errormetrics.ExecMissingParent)
processexecmetrics.MissingParentInc(parentId)
logger.GetLogger().WithField("processId", processId).WithField("parentId", parentId).Debug("Process missing parent")
} else {
if err == nil {
parent.RefInc()
}

Expand Down Expand Up @@ -79,10 +74,12 @@ func (msg *MsgExecveEventUnix) RetryInternal(ev notify.Event, timestamp uint64)
func (msg *MsgExecveEventUnix) Retry(internal *process.ProcessInternal, ev notify.Event) error {
var podInfo *tetragon.Pod

p := ev.GetProcess()
containerId := p.Docker
filename := p.Binary
args := p.Arguments
proc := ev.GetProcess()
parent := ev.GetParent()

containerId := proc.Docker
filename := proc.Binary
args := proc.Arguments
nspid := msg.Process.NSPID

if option.Config.EnableK8s && containerId != "" {
Expand All @@ -99,6 +96,20 @@ func (msg *MsgExecveEventUnix) Retry(internal *process.ProcessInternal, ev notif
// here.
internal.AddPodInfo(podInfo)
ev.SetProcess(internal.GetProcessCopy())

// Check we have a parent with exception for pid 1, note we do this last because we want
// to ensure the podInfo and process are set before returning any errors.
if proc.Pid.Value > 1 && parent == nil {
parentId := proc.ParentExecId
parent, err := process.Get(parentId)
if parent == nil {
return err
}
if strings.Contains(proc.Flags, "clone") == true {
parent.RefInc()
}
}

return nil
}

Expand All @@ -109,7 +120,9 @@ func (msg *MsgExecveEventUnix) HandleMessage() *tetragon.GetEventsResponse {
proc := process.AddExecEvent(&msg.MsgExecveEventUnix)
procEvent := GetProcessExec(proc)
ec := eventcache.Get()
if ec != nil && ec.Needed(procEvent.Process) {
if ec != nil &&
(ec.Needed(procEvent.Process) ||
(procEvent.Process.Pid.Value > 1 && ec.Needed(procEvent.Parent))) {
ec.Add(proc, procEvent, msg.MsgExecveEventUnix.Process.Ktime, msg)
} else {
procEvent.Process = proc.GetProcessCopy()
Expand Down Expand Up @@ -153,7 +166,6 @@ func GetProcessExit(event *MsgExitEventUnix) *tetragon.ProcessExit {

process, parent := process.GetParentProcessInternal(event.ProcessKey.Pid, event.ProcessKey.Ktime)
if process != nil {
process.RefDec()
tetragonProcess = process.UnsafeGetProcess()
} else {
tetragonProcess = &tetragon.Process{
Expand All @@ -162,7 +174,6 @@ func GetProcessExit(event *MsgExitEventUnix) *tetragon.ProcessExit {
}
}
if parent != nil {
parent.RefDec()
tetragonParent = parent.GetProcessCopy()
}

Expand All @@ -176,11 +187,17 @@ func GetProcessExit(event *MsgExitEventUnix) *tetragon.ProcessExit {
Status: code,
}
ec := eventcache.Get()
if ec != nil && ec.Needed(tetragonProcess) {
if ec != nil &&
(ec.Needed(tetragonProcess) ||
(tetragonProcess.Pid.Value > 1 && ec.Needed(tetragonParent))) {
ec.Add(process, tetragonEvent, event.ProcessKey.Ktime, event)
return nil
}
if parent != nil {
parent.RefDec()
}
if process != nil {
process.RefDec()
tetragonEvent.Process = process.GetProcessCopy()
}
return tetragonEvent
Expand All @@ -191,7 +208,26 @@ type MsgExitEventUnix struct {
}

func (msg *MsgExitEventUnix) RetryInternal(ev notify.Event, timestamp uint64) (*process.ProcessInternal, error) {
return eventcache.HandleGenericInternal(ev, timestamp)
p := ev.GetProcess()
internal, parent := process.GetParentProcessInternal(p.Pid.Value, timestamp)
var err error

if parent != nil {
ev.SetParent(parent.GetProcessCopy())
parent.RefDec()
} else {
errormetrics.ErrorTotalInc(errormetrics.EventCacheParentInfoFailed)
err = eventcache.ErrFailedToGetParentInfo
}

if internal != nil {
internal.RefDec()
} else {
errormetrics.ErrorTotalInc(errormetrics.EventCacheProcessInfoFailed)
err = eventcache.ErrFailedToGetProcessInfo
}

return internal, err
}

func (msg *MsgExitEventUnix) Retry(internal *process.ProcessInternal, ev notify.Event) error {
Expand Down
Loading