|
| 1 | +// |
| 2 | +// DISCLAIMER |
| 3 | +// |
| 4 | +// Copyright 2021 ArangoDB GmbH, Cologne, Germany |
| 5 | +// |
| 6 | +// Licensed under the Apache License, Version 2.0 (the "License"); |
| 7 | +// you may not use this file except in compliance with the License. |
| 8 | +// You may obtain a copy of the License at |
| 9 | +// |
| 10 | +// http://www.apache.org/licenses/LICENSE-2.0 |
| 11 | +// |
| 12 | +// Unless required by applicable law or agreed to in writing, software |
| 13 | +// distributed under the License is distributed on an "AS IS" BASIS, |
| 14 | +// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
| 15 | +// See the License for the specific language governing permissions and |
| 16 | +// limitations under the License. |
| 17 | +// |
| 18 | +// Copyright holder is ArangoDB GmbH, Cologne, Germany |
| 19 | +// |
| 20 | +// Author Adam Janikowski |
| 21 | +// |
| 22 | + |
| 23 | +package reconcile |
| 24 | + |
| 25 | +import ( |
| 26 | + "context" |
| 27 | + |
| 28 | + "github.com/arangodb/go-driver" |
| 29 | + "github.com/arangodb/kube-arangodb/pkg/util/arangod" |
| 30 | + |
| 31 | + "github.com/arangodb/kube-arangodb/pkg/util/errors" |
| 32 | + |
| 33 | + api "github.com/arangodb/kube-arangodb/pkg/apis/deployment/v1" |
| 34 | + "github.com/rs/zerolog" |
| 35 | +) |
| 36 | + |
| 37 | +func init() { |
| 38 | + registerAction(api.ActionTypeResignLeadership, newResignLeadershipAction) |
| 39 | +} |
| 40 | + |
| 41 | +// newResignLeadershipAction creates a new Action that implements the given |
| 42 | +// planned ResignLeadership action. |
| 43 | +func newResignLeadershipAction(log zerolog.Logger, action api.Action, actionCtx ActionContext) Action { |
| 44 | + a := &actionResignLeadership{} |
| 45 | + |
| 46 | + a.actionImpl = newActionImplDefRef(log, action, actionCtx, shutdownMemberTimeout) |
| 47 | + |
| 48 | + return a |
| 49 | +} |
| 50 | + |
| 51 | +// actionResignLeadership implements an ResignLeadershipAction. |
| 52 | +type actionResignLeadership struct { |
| 53 | + actionImpl |
| 54 | +} |
| 55 | + |
| 56 | +// Start performs the start of the ReasignLeadership process on DBServer. |
| 57 | +func (a *actionResignLeadership) Start(ctx context.Context) (bool, error) { |
| 58 | + log := a.log |
| 59 | + group := a.action.Group |
| 60 | + m, ok := a.actionCtx.GetMemberStatusByID(a.action.MemberID) |
| 61 | + if !ok { |
| 62 | + log.Error().Msg("No such member") |
| 63 | + return true, nil |
| 64 | + } |
| 65 | + |
| 66 | + if a.actionCtx.GetSpec().Mode.Get() != api.DeploymentModeCluster { |
| 67 | + log.Debug().Msg("Resign only allowed in cluster mode") |
| 68 | + return true, nil |
| 69 | + } |
| 70 | + |
| 71 | + client, err := a.actionCtx.GetDatabaseClient(ctx) |
| 72 | + if err != nil { |
| 73 | + log.Error().Err(err).Msgf("Unable to get client") |
| 74 | + return true, errors.WithStack(err) |
| 75 | + } |
| 76 | + |
| 77 | + switch group { |
| 78 | + case api.ServerGroupDBServers: |
| 79 | + cluster, err := client.Cluster(ctx) |
| 80 | + if err != nil { |
| 81 | + log.Error().Err(err).Msgf("Unable to get cluster client") |
| 82 | + return true, errors.WithStack(err) |
| 83 | + } |
| 84 | + |
| 85 | + var jobID string |
| 86 | + jobCtx := driver.WithJobIDResponse(ctx, &jobID) |
| 87 | + log.Debug().Msg("Temporary shutdown, resign leadership") |
| 88 | + if err := cluster.ResignServer(jobCtx, m.ID); err != nil { |
| 89 | + log.Debug().Err(err).Msg("Failed to resign server") |
| 90 | + return true, errors.WithStack(err) |
| 91 | + } |
| 92 | + |
| 93 | + m.CleanoutJobID = jobID |
| 94 | + |
| 95 | + if err := a.actionCtx.UpdateMember(m); err != nil { |
| 96 | + return true, errors.WithStack(err) |
| 97 | + } |
| 98 | + |
| 99 | + return false, nil |
| 100 | + default: |
| 101 | + return true, nil |
| 102 | + } |
| 103 | +} |
| 104 | + |
| 105 | +// CheckProgress checks if Job is completed. |
| 106 | +func (a *actionResignLeadership) CheckProgress(ctx context.Context) (bool, bool, error) { |
| 107 | + log := a.log |
| 108 | + |
| 109 | + m, ok := a.actionCtx.GetMemberStatusByID(a.action.MemberID) |
| 110 | + if !ok { |
| 111 | + log.Error().Msg("No such member") |
| 112 | + return true, false, nil |
| 113 | + } |
| 114 | + |
| 115 | + agency, err := a.actionCtx.GetAgency(ctx) |
| 116 | + if err != nil { |
| 117 | + log.Debug().Err(err).Msg("Failed to create agency client") |
| 118 | + return false, false, errors.WithStack(err) |
| 119 | + } |
| 120 | + |
| 121 | + c, err := a.actionCtx.GetDatabaseClient(ctx) |
| 122 | + if err != nil { |
| 123 | + log.Debug().Err(err).Msg("Failed to create member client") |
| 124 | + return false, false, errors.WithStack(err) |
| 125 | + } |
| 126 | + |
| 127 | + jobStatus, err := arangod.CleanoutServerJobStatus(ctx, m.CleanoutJobID, c, agency) |
| 128 | + if err != nil { |
| 129 | + if driver.IsNotFound(err) { |
| 130 | + log.Debug().Err(err).Msg("Job not found, but proceeding") |
| 131 | + return true, false, nil |
| 132 | + } |
| 133 | + log.Debug().Err(err).Msg("Failed to fetch job status") |
| 134 | + return false, false, errors.WithStack(err) |
| 135 | + } |
| 136 | + |
| 137 | + if jobStatus.IsFailed() { |
| 138 | + m.CleanoutJobID = "" |
| 139 | + if err := a.actionCtx.UpdateMember(m); err != nil { |
| 140 | + return false, false, errors.WithStack(err) |
| 141 | + } |
| 142 | + log.Error().Msg("Resign server job failed") |
| 143 | + return true, false, nil |
| 144 | + } |
| 145 | + |
| 146 | + if jobStatus.IsFinished() { |
| 147 | + m.CleanoutJobID = "" |
| 148 | + if err := a.actionCtx.UpdateMember(m); err != nil { |
| 149 | + return false, false, errors.WithStack(err) |
| 150 | + } |
| 151 | + return true, false, nil |
| 152 | + } |
| 153 | + |
| 154 | + return false, false, nil |
| 155 | +} |
0 commit comments