William Kurkian | ea86948 | 2019-04-09 15:16:11 -0400 | [diff] [blame] | 1 | package api |
| 2 | |
| 3 | import ( |
| 4 | "bytes" |
| 5 | "fmt" |
| 6 | "io" |
| 7 | "strconv" |
| 8 | "strings" |
| 9 | "time" |
| 10 | ) |
| 11 | |
| 12 | // AutopilotConfiguration is used for querying/setting the Autopilot configuration. |
| 13 | // Autopilot helps manage operator tasks related to Consul servers like removing |
| 14 | // failed servers from the Raft quorum. |
| 15 | type AutopilotConfiguration struct { |
| 16 | // CleanupDeadServers controls whether to remove dead servers from the Raft |
| 17 | // peer list when a new server joins |
| 18 | CleanupDeadServers bool |
| 19 | |
| 20 | // LastContactThreshold is the limit on the amount of time a server can go |
| 21 | // without leader contact before being considered unhealthy. |
| 22 | LastContactThreshold *ReadableDuration |
| 23 | |
| 24 | // MaxTrailingLogs is the amount of entries in the Raft Log that a server can |
| 25 | // be behind before being considered unhealthy. |
| 26 | MaxTrailingLogs uint64 |
| 27 | |
| 28 | // ServerStabilizationTime is the minimum amount of time a server must be |
| 29 | // in a stable, healthy state before it can be added to the cluster. Only |
| 30 | // applicable with Raft protocol version 3 or higher. |
| 31 | ServerStabilizationTime *ReadableDuration |
| 32 | |
| 33 | // (Enterprise-only) RedundancyZoneTag is the node tag to use for separating |
| 34 | // servers into zones for redundancy. If left blank, this feature will be disabled. |
| 35 | RedundancyZoneTag string |
| 36 | |
| 37 | // (Enterprise-only) DisableUpgradeMigration will disable Autopilot's upgrade migration |
| 38 | // strategy of waiting until enough newer-versioned servers have been added to the |
| 39 | // cluster before promoting them to voters. |
| 40 | DisableUpgradeMigration bool |
| 41 | |
| 42 | // (Enterprise-only) UpgradeVersionTag is the node tag to use for version info when |
| 43 | // performing upgrade migrations. If left blank, the Consul version will be used. |
| 44 | UpgradeVersionTag string |
| 45 | |
| 46 | // CreateIndex holds the index corresponding the creation of this configuration. |
| 47 | // This is a read-only field. |
| 48 | CreateIndex uint64 |
| 49 | |
| 50 | // ModifyIndex will be set to the index of the last update when retrieving the |
| 51 | // Autopilot configuration. Resubmitting a configuration with |
| 52 | // AutopilotCASConfiguration will perform a check-and-set operation which ensures |
| 53 | // there hasn't been a subsequent update since the configuration was retrieved. |
| 54 | ModifyIndex uint64 |
| 55 | } |
| 56 | |
| 57 | // ServerHealth is the health (from the leader's point of view) of a server. |
| 58 | type ServerHealth struct { |
| 59 | // ID is the raft ID of the server. |
| 60 | ID string |
| 61 | |
| 62 | // Name is the node name of the server. |
| 63 | Name string |
| 64 | |
| 65 | // Address is the address of the server. |
| 66 | Address string |
| 67 | |
| 68 | // The status of the SerfHealth check for the server. |
| 69 | SerfStatus string |
| 70 | |
| 71 | // Version is the Consul version of the server. |
| 72 | Version string |
| 73 | |
| 74 | // Leader is whether this server is currently the leader. |
| 75 | Leader bool |
| 76 | |
| 77 | // LastContact is the time since this node's last contact with the leader. |
| 78 | LastContact *ReadableDuration |
| 79 | |
| 80 | // LastTerm is the highest leader term this server has a record of in its Raft log. |
| 81 | LastTerm uint64 |
| 82 | |
| 83 | // LastIndex is the last log index this server has a record of in its Raft log. |
| 84 | LastIndex uint64 |
| 85 | |
| 86 | // Healthy is whether or not the server is healthy according to the current |
| 87 | // Autopilot config. |
| 88 | Healthy bool |
| 89 | |
| 90 | // Voter is whether this is a voting server. |
| 91 | Voter bool |
| 92 | |
| 93 | // StableSince is the last time this server's Healthy value changed. |
| 94 | StableSince time.Time |
| 95 | } |
| 96 | |
| 97 | // OperatorHealthReply is a representation of the overall health of the cluster |
| 98 | type OperatorHealthReply struct { |
| 99 | // Healthy is true if all the servers in the cluster are healthy. |
| 100 | Healthy bool |
| 101 | |
| 102 | // FailureTolerance is the number of healthy servers that could be lost without |
| 103 | // an outage occurring. |
| 104 | FailureTolerance int |
| 105 | |
| 106 | // Servers holds the health of each server. |
| 107 | Servers []ServerHealth |
| 108 | } |
| 109 | |
| 110 | // ReadableDuration is a duration type that is serialized to JSON in human readable format. |
| 111 | type ReadableDuration time.Duration |
| 112 | |
| 113 | func NewReadableDuration(dur time.Duration) *ReadableDuration { |
| 114 | d := ReadableDuration(dur) |
| 115 | return &d |
| 116 | } |
| 117 | |
| 118 | func (d *ReadableDuration) String() string { |
| 119 | return d.Duration().String() |
| 120 | } |
| 121 | |
| 122 | func (d *ReadableDuration) Duration() time.Duration { |
| 123 | if d == nil { |
| 124 | return time.Duration(0) |
| 125 | } |
| 126 | return time.Duration(*d) |
| 127 | } |
| 128 | |
| 129 | func (d *ReadableDuration) MarshalJSON() ([]byte, error) { |
| 130 | return []byte(fmt.Sprintf(`"%s"`, d.Duration().String())), nil |
| 131 | } |
| 132 | |
| 133 | func (d *ReadableDuration) UnmarshalJSON(raw []byte) error { |
| 134 | if d == nil { |
| 135 | return fmt.Errorf("cannot unmarshal to nil pointer") |
| 136 | } |
| 137 | |
| 138 | str := string(raw) |
| 139 | if len(str) < 2 || str[0] != '"' || str[len(str)-1] != '"' { |
| 140 | return fmt.Errorf("must be enclosed with quotes: %s", str) |
| 141 | } |
| 142 | dur, err := time.ParseDuration(str[1 : len(str)-1]) |
| 143 | if err != nil { |
| 144 | return err |
| 145 | } |
| 146 | *d = ReadableDuration(dur) |
| 147 | return nil |
| 148 | } |
| 149 | |
| 150 | // AutopilotGetConfiguration is used to query the current Autopilot configuration. |
| 151 | func (op *Operator) AutopilotGetConfiguration(q *QueryOptions) (*AutopilotConfiguration, error) { |
| 152 | r := op.c.newRequest("GET", "/v1/operator/autopilot/configuration") |
| 153 | r.setQueryOptions(q) |
| 154 | _, resp, err := requireOK(op.c.doRequest(r)) |
| 155 | if err != nil { |
| 156 | return nil, err |
| 157 | } |
| 158 | defer resp.Body.Close() |
| 159 | |
| 160 | var out AutopilotConfiguration |
| 161 | if err := decodeBody(resp, &out); err != nil { |
| 162 | return nil, err |
| 163 | } |
| 164 | |
| 165 | return &out, nil |
| 166 | } |
| 167 | |
| 168 | // AutopilotSetConfiguration is used to set the current Autopilot configuration. |
| 169 | func (op *Operator) AutopilotSetConfiguration(conf *AutopilotConfiguration, q *WriteOptions) error { |
| 170 | r := op.c.newRequest("PUT", "/v1/operator/autopilot/configuration") |
| 171 | r.setWriteOptions(q) |
| 172 | r.obj = conf |
| 173 | _, resp, err := requireOK(op.c.doRequest(r)) |
| 174 | if err != nil { |
| 175 | return err |
| 176 | } |
| 177 | resp.Body.Close() |
| 178 | return nil |
| 179 | } |
| 180 | |
| 181 | // AutopilotCASConfiguration is used to perform a Check-And-Set update on the |
| 182 | // Autopilot configuration. The ModifyIndex value will be respected. Returns |
| 183 | // true on success or false on failures. |
| 184 | func (op *Operator) AutopilotCASConfiguration(conf *AutopilotConfiguration, q *WriteOptions) (bool, error) { |
| 185 | r := op.c.newRequest("PUT", "/v1/operator/autopilot/configuration") |
| 186 | r.setWriteOptions(q) |
| 187 | r.params.Set("cas", strconv.FormatUint(conf.ModifyIndex, 10)) |
| 188 | r.obj = conf |
| 189 | _, resp, err := requireOK(op.c.doRequest(r)) |
| 190 | if err != nil { |
| 191 | return false, err |
| 192 | } |
| 193 | defer resp.Body.Close() |
| 194 | |
| 195 | var buf bytes.Buffer |
| 196 | if _, err := io.Copy(&buf, resp.Body); err != nil { |
| 197 | return false, fmt.Errorf("Failed to read response: %v", err) |
| 198 | } |
| 199 | res := strings.Contains(buf.String(), "true") |
| 200 | |
| 201 | return res, nil |
| 202 | } |
| 203 | |
| 204 | // AutopilotServerHealth |
| 205 | func (op *Operator) AutopilotServerHealth(q *QueryOptions) (*OperatorHealthReply, error) { |
| 206 | r := op.c.newRequest("GET", "/v1/operator/autopilot/health") |
| 207 | r.setQueryOptions(q) |
| 208 | _, resp, err := requireOK(op.c.doRequest(r)) |
| 209 | if err != nil { |
| 210 | return nil, err |
| 211 | } |
| 212 | defer resp.Body.Close() |
| 213 | |
| 214 | var out OperatorHealthReply |
| 215 | if err := decodeBody(resp, &out); err != nil { |
| 216 | return nil, err |
| 217 | } |
| 218 | return &out, nil |
| 219 | } |