github.com/1aal/kubeblocks@v0.0.0-20231107070852-e1c03e598921/pkg/lorry/operations/replica/checkstatus.go (about)

     1  /*
     2  Copyright (C) 2022-2023 ApeCloud Co., Ltd
     3  
     4  This file is part of KubeBlocks project
     5  
     6  This program is free software: you can redistribute it and/or modify
     7  it under the terms of the GNU Affero General Public License as published by
     8  the Free Software Foundation, either version 3 of the License, or
     9  (at your option) any later version.
    10  
    11  This program is distributed in the hope that it will be useful
    12  but WITHOUT ANY WARRANTY; without even the implied warranty of
    13  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
    14  GNU Affero General Public License for more details.
    15  
    16  You should have received a copy of the GNU Affero General Public License
    17  along with this program.  If not, see <http://www.gnu.org/licenses/>.
    18  */
    19  
    20  package replica
    21  
    22  import (
    23  	"context"
    24  
    25  	"github.com/go-logr/logr"
    26  	"github.com/pkg/errors"
    27  	"github.com/spf13/viper"
    28  	ctrl "sigs.k8s.io/controller-runtime"
    29  
    30  	"github.com/1aal/kubeblocks/pkg/lorry/dcs"
    31  	"github.com/1aal/kubeblocks/pkg/lorry/engines"
    32  	"github.com/1aal/kubeblocks/pkg/lorry/engines/register"
    33  	"github.com/1aal/kubeblocks/pkg/lorry/operations"
    34  	"github.com/1aal/kubeblocks/pkg/lorry/util"
    35  )
    36  
    37  type CheckStatus struct {
    38  	operations.Base
    39  	dcsStore                   dcs.DCS
    40  	dbManager                  engines.DBManager
    41  	checkFailedCount           int
    42  	failedEventReportFrequency int
    43  	logger                     logr.Logger
    44  }
    45  
    46  var checkstatus operations.Operation = &CheckStatus{}
    47  
    48  func init() {
    49  	err := operations.Register("checkstatus", checkstatus)
    50  	if err != nil {
    51  		panic(err.Error())
    52  	}
    53  }
    54  
    55  func (s *CheckStatus) Init(ctx context.Context) error {
    56  	s.dcsStore = dcs.GetStore()
    57  	if s.dcsStore == nil {
    58  		return errors.New("dcs store init failed")
    59  	}
    60  
    61  	dbManager, err := register.GetDBManager()
    62  	if err != nil {
    63  		return errors.Wrap(err, "get manager failed")
    64  	}
    65  	s.dbManager = dbManager
    66  
    67  	s.failedEventReportFrequency = viper.GetInt("KB_FAILED_EVENT_REPORT_FREQUENCY")
    68  	if s.failedEventReportFrequency < 300 {
    69  		s.failedEventReportFrequency = 300
    70  	} else if s.failedEventReportFrequency > 3600 {
    71  		s.failedEventReportFrequency = 3600
    72  	}
    73  
    74  	s.logger = ctrl.Log.WithName("checkstatus")
    75  	return nil
    76  }
    77  
    78  func (s *CheckStatus) IsReadonly(ctx context.Context) bool {
    79  	return true
    80  }
    81  
    82  func (s *CheckStatus) Do(ctx context.Context, req *operations.OpsRequest) (*operations.OpsResponse, error) {
    83  	resp := &operations.OpsResponse{
    84  		Data: map[string]any{},
    85  	}
    86  	resp.Data["operation"] = util.CheckStatusOperation
    87  	var message string
    88  
    89  	k8sStore := s.dcsStore.(*dcs.KubernetesStore)
    90  	cluster := k8sStore.GetClusterFromCache()
    91  	isHealthy := s.dbManager.IsCurrentMemberHealthy(ctx, cluster)
    92  	if !isHealthy {
    93  		message = "status check failed"
    94  		s.logger.Info(message)
    95  		resp.Data["event"] = util.OperationFailed
    96  		resp.Data["message"] = message
    97  		var err error
    98  		if s.checkFailedCount%s.failedEventReportFrequency == 0 {
    99  			s.logger.Info("status checks failed continuously", "times", s.checkFailedCount)
   100  			err = util.SentEventForProbe(ctx, resp.Data)
   101  		}
   102  		s.checkFailedCount++
   103  		return resp, err
   104  	}
   105  	s.checkFailedCount = 0
   106  	resp.Data["event"] = util.OperationSuccess
   107  	return resp, nil
   108  }