Add nodes and cluster settings support (#9)

This commit is contained in:
T. von Dein
2026-05-05 12:06:34 +02:00
parent a169d23b90
commit fcfeaf66a9
7 changed files with 265 additions and 335 deletions

View File

@@ -18,15 +18,14 @@ package es
import (
"context"
"encoding/json"
"errors"
"fmt"
"log"
"log/slog"
"regexp"
"codeberg.org/scip/esctl/pkg/cfg"
"github.com/elastic/go-elasticsearch/v9/typedapi/cluster/health"
"github.com/elastic/go-elasticsearch/v9/typedapi/types"
"github.com/urfave/cli/v3"
)
const (
@@ -75,7 +74,7 @@ func ClusterCompare(conf *cfg.Config, leader, follower string) error {
return nil
}
func List(conf *cfg.Config) error {
func ClusterList(conf *cfg.Config) error {
table := NewTable(2, len(conf.Clusters))
table.Addheaders("cluster", "uri")
@@ -91,7 +90,7 @@ func List(conf *cfg.Config) error {
return nil
}
func Status(conf *cfg.Config) error {
func ClusterStatus(conf *cfg.Config) error {
clusters := []string{}
if conf.All {
@@ -113,7 +112,7 @@ func Status(conf *cfg.Config) error {
Header("accept", "application/json").
Do(context.Background())
if err != nil {
log.Fatalf("Error getting health: %s", err)
return fmt.Errorf("failed to getcluster health: %s", err)
}
slog.Debug("ES result", "cluster health", res)
@@ -135,340 +134,79 @@ func Status(conf *cfg.Config) error {
return nil
}
// look for indicies only on leader
func findIndicesOnlyOnMaster(conf *cfg.Config, indices ClusterIndices, leader, follower string) {
exclude := regexp.MustCompile(DefaultExclude)
if conf.Exclude != "" {
exclude = regexp.MustCompile(conf.Exclude)
}
indexOnlyOnLeader := map[string]*types.IndicesRecord{}
for name, index := range indices[leader] {
if exclude.MatchString(name) {
continue
}
_, followerHasIt := indices[follower][name]
if !followerHasIt {
// fetch index details
res, err := conf.Clusters[leader].ES.Indices.Get(name).
Header("content-type", "application/json").
Header("accept", "application/json").
Do(context.Background())
if err != nil {
continue // ignore it then
}
_, defined := res[name]
if !defined {
// json response map didn't contain the index
continue
}
isWritable := false
for _, alias := range res[name].Aliases {
if *alias.IsWriteIndex {
isWritable = true
break
}
}
if isWritable {
// ignore index if associated alias index is writing
continue
}
indexOnlyOnLeader[name] = index
}
}
idx := 0
table := NewTable(3, len(indexOnlyOnLeader))
table.Addheaders("index only on leader", "size", "docscount")
for name, index := range indexOnlyOnLeader {
name := Colorize("red", name)
table.entries[idx] = []string{name, *index.DatasetSize, *index.DocsCount}
idx++
}
table.Sort()
table.PrintMarkdown()
}
func checkClusterFollower(conf *cfg.Config, leader string) bool {
stats, err := conf.Clusters[leader].ES.Ccr.Stats().
func ClusterSettingsList(conf *cfg.Config) error {
res, err := conf.DefaultCluster.ES.Cluster.GetSettings().
Header("content-type", "application/json").
Header("accept", "application/json").
Do(context.Background())
if err != nil {
fmt.Printf("failed to get ccr stats from %s: %s", leader, err)
return false
return fmt.Errorf("failed to get cluster settings: %s", err)
}
if len(stats.AutoFollowStats.AutoFollowedClusters) == 0 {
// is not following anyone
return true
}
table := NewTable(2, 0)
table.Addheaders("setting", "value")
entries := [][]string{}
if stats.AutoFollowStats.AutoFollowedClusters[0].ClusterName != "" {
fmt.Println("leader/follower attribution is invalid, reverse cluster attribution and retry")
return false
}
for topic, val := range res.Persistent {
data := map[string]map[string]any{}
return true
}
err := json.Unmarshal(val, &data)
// checks if both clusters are green
func checkClusterStatus(conf *cfg.Config, leader, follower string) bool {
status := map[string]*health.Response{}
for _, cluster := range []string{leader, follower} {
st, err := conf.Clusters[leader].ES.Cluster.Health().
Header("content-type", "application/json").
Header("accept", "application/json").
Do(context.Background())
if err != nil {
fmt.Printf("failed to get health from %s: %s", cluster, err)
return false
return fmt.Errorf("failed to unmarshall setting for topic %s: %s", topic, err)
}
status[cluster] = st
for key, settings := range data {
for setting, value := range settings {
entries = append(entries, []string{
fmt.Sprintf("%s.%s.%s", topic, key, setting),
fmt.Sprintf("%v", value),
})
}
}
}
table := NewTable(3, 5)
table.Addheaders("setting", "leader:"+leader, "follower:"+follower)
table.entries = [][]string{
{"Cluster Name",
Colorize(*&status[leader].Status.Name, status[leader].ClusterName),
Colorize(*&status[follower].Status.Name, status[follower].ClusterName),
},
{"Active Shards",
fmt.Sprintf("%d", status[leader].ActiveShards),
fmt.Sprintf("%d", status[follower].ActiveShards),
},
{"Active Primary Shards",
fmt.Sprintf("%d", status[leader].ActivePrimaryShards),
fmt.Sprintf("%d", status[follower].ActivePrimaryShards),
},
{"Indicies",
fmt.Sprintf("%d", len(status[leader].Indices)),
fmt.Sprintf("%d", len(status[follower].Indices)),
},
{"Nodes",
fmt.Sprintf("%d", status[leader].NumberOfNodes),
fmt.Sprintf("%d", status[follower].NumberOfNodes),
},
}
table.entries = entries
table.Sort()
table.PrintMarkdown()
if status[leader].Status.Name == "green" && status[follower].Status.Name == "green" {
return true
}
return false
return nil
}
// finds indices on both clusters which have ilm errors
func findIlmErrors(conf *cfg.Config, leader, follower string) bool {
failed := map[string]map[string]string{}
func ClusterSettingsSet(conf *cfg.Config, args cli.Args) error {
put := conf.Clusters[conf.CurrentCluster].ES.Cluster.PutSettings()
for _, cluster := range []string{leader, follower} {
ilm, err := conf.Clusters[cluster].ES.Ilm.ExplainLifecycle("_all").
OnlyManaged(true).
Header("content-type", "application/json").
Header("accept", "application/json").
Do(context.Background())
if err != nil {
fmt.Printf("failed to get ilm status from %s: %s", cluster, err)
return false
}
for _, arg := range args.Slice() {
setting, value := splitArg(arg)
failed[cluster] = map[string]string{}
for name, ilmstate := range ilm.Indices {
count := ilmstate.(*types.LifecycleExplainManaged).FailedStepRetryCount
if count != nil && *count > 0 {
failed[cluster][name] = fmt.Sprintf("%d", *count)
}
}
}
if len(failed[leader]) == 0 && len(failed[follower]) == 0 {
return true
}
for idx, cluster := range []string{leader, follower} {
which := "leader"
if idx > 0 {
which = "follower"
}
if len(failed[cluster]) > 0 {
idx := 0
table := NewTable(2, len(failed[cluster]))
table.Addheaders("ilm errors on "+which, "errors")
for name, count := range failed[cluster] {
table.entries[idx] = []string{name, count}
idx++
}
table.Sort()
table.PrintMarkdown()
}
}
return false
}
// find unsynchronized indicies only present on leader
func findIndicesOnlyOnLeader(conf *cfg.Config, indices ClusterIndices, leader, follower string) bool {
exclude := regexp.MustCompile(DefaultExclude)
if conf.Exclude != "" {
exclude = regexp.MustCompile(conf.Exclude)
}
indexOnlyOnLeader := map[string]*types.IndicesRecord{}
for name, index := range indices[leader] {
if exclude.MatchString(name) {
continue
}
_, followerHasIt := indices[follower][name]
if !followerHasIt {
// fetch index details
res, err := conf.Clusters[leader].ES.Indices.Get(name).
Header("content-type", "application/json").
Header("accept", "application/json").
Do(context.Background())
switch {
case conf.Transient:
message, err := json.Marshal(value)
if err != nil {
continue // ignore it then
return fmt.Errorf("failed to marshall transient value <%v> to valid JSON: %s", value, err)
}
_, defined := res[name]
if !defined {
// json response map didn't contain the index
continue
}
isWritable := false
for _, alias := range res[name].Aliases {
if *alias.IsWriteIndex {
isWritable = true
break
}
}
if isWritable {
// ignore index if associated alias index is writing
continue
}
indexOnlyOnLeader[name] = index
}
}
if len(indexOnlyOnLeader) == 0 {
return true
}
idx := 0
table := NewTable(3, len(indexOnlyOnLeader))
table.Addheaders("index only on leader", "size", "docscount")
for name, index := range indexOnlyOnLeader {
name := Colorize("red", name)
table.entries[idx] = []string{name, *index.DatasetSize, *index.DocsCount}
idx++
}
table.Sort()
table.PrintMarkdown()
return false
}
// find indices only present on follower
func findOrphanedIndices(conf *cfg.Config, indices ClusterIndices, leader, follower string) bool {
orphaned := map[string]*types.IndicesRecord{}
for name, index := range indices[follower] {
_, leaderHasIt := indices[leader][name]
if !leaderHasIt {
// fetch index details
res, err := conf.Clusters[follower].ES.Indices.Get(name).
Header("content-type", "application/json").
Header("accept", "application/json").
Do(context.Background())
put.AddTransient(setting, message)
case conf.Persistent:
fallthrough
default:
message, err := json.Marshal(value)
if err != nil {
continue // ignore it then
return fmt.Errorf("failed to marshall persistent value <%v> to valid JSON: %s", value, err)
}
_, defined := res[name]
if !defined {
// json response map didn't contain the index
continue
}
orphaned[name] = index
put.AddPersistent(setting, message)
}
}
if len(orphaned) == 0 {
return true
_, err := put.
Header("content-type", "application/json").
Header("accept", "application/json").
Do(context.Background())
if err != nil {
return fmt.Errorf("failed to set settings: %s", err)
}
idx := 0
table := NewTable(3, len(orphaned))
table.Addheaders("orphaned index on follower", "size", "docscount")
for name, index := range orphaned {
name := Colorize("red", name)
table.entries[idx] = []string{name, *index.DatasetSize, *index.DocsCount}
idx++
}
table.Sort()
table.PrintMarkdown()
return false
}
// find red indices on follower
func findFailedFollowerIndices(conf *cfg.Config, indices ClusterIndices, follower string) bool {
red := map[string]*types.IndicesRecord{}
for name, index := range indices[follower] {
if *index.Health == "red" {
red[name] = index
}
}
if len(red) == 0 {
return true
}
idx := 0
table := NewTable(3, len(red))
table.Addheaders("red index on follower", "size", "docscount")
for name, index := range red {
name := Colorize("red", name)
table.entries[idx] = []string{name, *index.DatasetSize, *index.DocsCount}
idx++
}
table.Sort()
table.PrintMarkdown()
return false
return nil
}

55
pkg/es/node.go Normal file
View File

@@ -0,0 +1,55 @@
/*
Copyright © 2026 Thomas von Dein
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
package es
import (
"context"
"fmt"
"log/slog"
"codeberg.org/scip/esctl/pkg/cfg"
)
func NodeList(conf *cfg.Config) error {
// get nodes
nodes, err := conf.DefaultCluster.ES.Cat.Nodes().Do(context.Background())
if err != nil {
return fmt.Errorf("Error getting nodes: %s", err)
}
slog.Debug("ES result", "nodes", nodes)
table := NewTable(7, len(nodes))
table.Addheaders("name", "ip", "load1m", "load5m", "load15m", "ram %", "heap %")
for idx, node := range nodes {
table.entries[idx] = []string{
*node.Name,
*node.Ip,
*node.Load1M,
*node.Load5M,
*node.Load15M,
node.RamPercent.(string),
node.HeapPercent.(string),
}
}
table.Sort()
table.PrintMarkdown()
return nil
}