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

[action] [PR:261] [active-standby] Fix the oscillation logic (#261) #269

Merged
merged 1 commit into from
Aug 23, 2024
Merged

[action] [PR:261] [active-standby] Fix the oscillation logic (#261) #269

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
31 changes: 28 additions & 3 deletions src/link_manager/LinkManagerStateMachineActiveStandby.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -61,8 +61,6 @@ ActiveStandbyStateMachine::ActiveStandbyStateMachine(
mMuxStateMachine.setWaitStateCause(mux_state::WaitState::WaitStateCause::SwssUpdate);
mMuxPortPtr->setMuxLinkmgrState(mLabel);
initializeTransitionFunctionTable();

mOscillationTimer.expires_from_now(boost::posix_time::seconds(1));
}

//
Expand Down Expand Up @@ -476,12 +474,16 @@ void ActiveStandbyStateMachine::handleStateChange(LinkProberEvent &event, link_p
mMuxPortPtr->postLinkProberMetricsEvent(link_manager::ActiveStandbyStateMachine::LinkProberMetrics::LinkProberActiveStart);

mStandbyUnknownUpCount = 0;

tryCancelOscillationTimerIfAlive();
}

if (state == link_prober::LinkProberState::Label::Standby) {
mMuxPortPtr->postLinkProberMetricsEvent(link_manager::ActiveStandbyStateMachine::LinkProberMetrics::LinkProberStandbyStart);

mActiveUnknownUpCount = 0;

tryCancelOscillationTimerIfAlive();
}

CompositeState nextState = mCompositeState;
Expand Down Expand Up @@ -542,6 +544,10 @@ void ActiveStandbyStateMachine::handleStateChange(MuxStateEvent &event, mux_stat
mStandbyUnknownUpCount = 0;
}

if (state == mux_state::MuxState::Label::Standby) {
tryCancelOscillationTimerIfAlive();
}

updateMuxLinkmgrState();
}

Expand Down Expand Up @@ -582,6 +588,8 @@ void ActiveStandbyStateMachine::handleStateChange(LinkStateEvent &event, link_st
mWaitActiveUpCount = 0;
mWaitStandbyUpBackoffFactor = 1;
mUnknownActiveUpBackoffFactor = 1;

tryCancelOscillationTimerIfAlive();
} else {
mStateTransitionHandler[ps(nextState)][ms(nextState)][ls(nextState)](nextState);
}
Expand Down Expand Up @@ -1064,6 +1072,8 @@ void ActiveStandbyStateMachine::handleMuxWaitTimeout(boost::system::error_code e
void ActiveStandbyStateMachine::startOscillationTimer()
{
// Note: This timer is started when Mux state is active and link prober is in wait state.
MUXLOGINFO(boost::format("%s: start the oscillation timer") % mMuxPortConfig.getPortName());
mOscillationTimerAlive = true;
mOscillationTimer.expires_from_now(boost::posix_time::seconds(
mMuxPortConfig.getOscillationInterval_sec()
));
Expand All @@ -1074,6 +1084,20 @@ void ActiveStandbyStateMachine::startOscillationTimer()
)));
}

//
// ---> tryCancelOscillationTimerIfAlive();
//
// cancel the oscillation timer if it is alive
//
void ActiveStandbyStateMachine::tryCancelOscillationTimerIfAlive()
{
if (mOscillationTimerAlive) {
MUXLOGINFO(boost::format("%s: cancel the oscillation timer") % mMuxPortConfig.getPortName());
mOscillationTimerAlive = false;
mOscillationTimer.cancel();
}
}

//
// ---> handleOscillationTimeout(boost::system::error_code errorCode);
//
Expand All @@ -1083,6 +1107,7 @@ void ActiveStandbyStateMachine::handleOscillationTimeout(boost::system::error_co
{
MUXLOGDEBUG(mMuxPortConfig.getPortName());

mOscillationTimerAlive = false;
if (mMuxPortConfig.getIfOscillationEnabled() &&
errorCode == boost::system::errc::success &&
ps(mCompositeState) == link_prober::LinkProberState::Label::Wait &&
Expand Down Expand Up @@ -1318,7 +1343,7 @@ void ActiveStandbyStateMachine::LinkProberWaitMuxActiveLinkUpTransitionFunction(
mSuspendTxFnPtr(mMuxPortConfig.getLinkWaitTimeout_msec());
}

if (mOscillationTimer.expires_at() < boost::posix_time::microsec_clock::local_time()) {
if (!mOscillationTimerAlive) {
startOscillationTimer();
}
}
Expand Down
10 changes: 10 additions & 0 deletions src/link_manager/LinkManagerStateMachineActiveStandby.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -464,6 +464,15 @@ class ActiveStandbyStateMachine: public LinkManagerStateMachineBase,
*/
void startOscillationTimer();

/**
*@method tryCancelOscillationTimerIfAlive
*
*@brief cancel the oscillation timer if it is alive
*
*@return none
*/
inline void tryCancelOscillationTimerIfAlive();

/**
*@method handleOscillationTimeout
*
Expand Down Expand Up @@ -859,6 +868,7 @@ class ActiveStandbyStateMachine: public LinkManagerStateMachineBase,
boost::asio::deadline_timer mDeadlineTimer;
boost::asio::deadline_timer mWaitTimer;
boost::asio::deadline_timer mOscillationTimer;
bool mOscillationTimerAlive = false;

boost::function<void ()> mInitializeProberFnPtr;
boost::function<void ()> mStartProbingFnPtr;
Expand Down
39 changes: 37 additions & 2 deletions test/LinkManagerStateMachineTest.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,6 +21,9 @@
* Author: Tamer Ahmed
*/

#include <boost/date_time/posix_time/posix_time.hpp>
#include <boost/thread/thread.hpp>

#include "LinkManagerStateMachineTest.h"
#include "link_prober/LinkProberStateMachineBase.h"
#include "common/MuxLogger.h"
Expand Down Expand Up @@ -50,7 +53,7 @@ LinkManagerStateMachineTest::LinkManagerStateMachineTest() :
mMuxConfig.setPositiveStateChangeRetryCount(mPositiveUpdateCount);
mMuxConfig.setMuxStateChangeRetryCount(mPositiveUpdateCount);
mMuxConfig.setLinkStateChangeRetryCount(mPositiveUpdateCount);
mMuxConfig.setOscillationInterval_sec(1,true);
mMuxConfig.setOscillationInterval_sec(2, true);
}

void LinkManagerStateMachineTest::runIoService(uint32_t count)
Expand Down Expand Up @@ -1468,7 +1471,7 @@ TEST_F(LinkManagerStateMachineTest, TimedOscillation)
handleMuxState("active", 3);
VALIDATE_STATE(Wait, Active, Up);

runIoService(2);
runIoService(1);
VALIDATE_STATE(Wait, Wait, Up);

handleProbeMuxState("active", 3);
Expand All @@ -1481,6 +1484,38 @@ TEST_F(LinkManagerStateMachineTest, TimedOscillation)
mMuxConfig.setTimeoutIpv4_msec(10);
}

TEST_F(LinkManagerStateMachineTest, TimedOscillationMuxProbeStandbyCancel)
{
setMuxStandby();

// set icmp timeout to be 500ms otherwise it will probe mux state endlessly and get no chance to do timed oscillation
mMuxConfig.setTimeoutIpv4_msec(500);

postLinkProberEvent(link_prober::LinkProberState::Unknown, 2);
VALIDATE_STATE(Wait, Wait, Up);
EXPECT_EQ(mDbInterfacePtr->mSetMuxStateInvokeCount, 1);
EXPECT_EQ(mDbInterfacePtr->mLastPostedSwitchCause, link_manager::ActiveStandbyStateMachine::SwitchCause::PeerHeartbeatMissing);

// swss notification
handleMuxState("active", 3);
VALIDATE_STATE(Wait, Active, Up);

runIoService(1);
VALIDATE_STATE(Wait, Wait, Up);

handleProbeMuxState("standby", 3);
VALIDATE_STATE(Wait, Standby, Up);

boost::this_thread::sleep(boost::posix_time::seconds(2));

// 3 mux probe active notifiations + 1 oscillation timeout + 1 mux probe timeout
handleProbeMuxState("active", 5);
VALIDATE_STATE(Wait, Active, Up);
EXPECT_EQ(mDbInterfacePtr->mLastPostedSwitchCause, link_manager::ActiveStandbyStateMachine::SwitchCause::PeerHeartbeatMissing);

mMuxConfig.setTimeoutIpv4_msec(10);
}

TEST_F(LinkManagerStateMachineTest, OrchagentRollback)
{
setMuxStandby();
Expand Down
Loading