Hello everyone, i’m attempting to put a cluster to work using ec2 instances.
I launched 3 machines, and would like to put one zero in each one and 2 alphas, however, i’m facing some weird behaviour i’m not understanding. Questions following below.
10.30.0.191 (zero (leader))
10.30.0.180 (zero and alpha)
10.30.0.26 (zero and alpha (leader))
Zero → 6080/State:
{
"counter": "1034",
"groups": {
"1": {
"members": {
"1": {
"id": "1",
"groupId": 1,
"addr": "10.30.0.180:7080",
"lastUpdate": "1607356083"
},
"2": {
"id": "2",
"groupId": 1,
"addr": "10.30.0.26:7080",
"leader": true,
"lastUpdate": "1607356282"
}
},
"tablets": {
"Album": {
"groupId": 1,
"predicate": "Album"
},
"Song": {
"groupId": 1,
"predicate": "Song"
},
"dgraph.acl.rule": {
"groupId": 1,
"predicate": "dgraph.acl.rule"
},
"dgraph.graphql.schema": {
"groupId": 1,
"predicate": "dgraph.graphql.schema"
},
"dgraph.graphql.xid": {
"groupId": 1,
"predicate": "dgraph.graphql.xid"
},
"dgraph.password": {
"groupId": 1,
"predicate": "dgraph.password"
},
"dgraph.rule.permission": {
"groupId": 1,
"predicate": "dgraph.rule.permission"
},
"dgraph.rule.predicate": {
"groupId": 1,
"predicate": "dgraph.rule.predicate"
},
"dgraph.type": {
"groupId": 1,
"predicate": "dgraph.type"
},
"dgraph.user.group": {
"groupId": 1,
"predicate": "dgraph.user.group"
},
"dgraph.xid": {
"groupId": 1,
"predicate": "dgraph.xid"
},
"type": {
"groupId": 1,
"predicate": "type"
}
},
"checksum": "8242878090081052337"
}
},
"zeros": {
"1": {
"id": "1",
"addr": "localhost:5080"
},
"2": {
"id": "2",
"addr": "10.30.0.191:5080",
"leader": true
},
"3": {
"id": "3",
"addr": "10.30.0.180:5080"
}
},
"maxLeaseId": "8390000",
"maxTxnTs": "150000",
"maxRaftId": "2",
"cid": "08104978-e707-487e-8090-22201fb88506",
"license": {
"maxNodes": "18446744073709551615",
"expiryTs": "1609105755",
"enabled": true
}
}
- Why zero ID:1 is localhost:5080?
So I just requested /health?all alphas: 10.30.0.26:7080 and 10.30.0.180:7080
//10.30.0.180:7080
[
{
"instance": "zero",
"address": "10.30.0.191:5080",
"status": "healthy",
"group": "0",
"version": "v20.07.2",
"uptime": 6013,
"lastEcho": 1607358147
},
{
"instance": "alpha",
"address": "10.30.0.26:7080",
"status": "healthy",
"group": "1",
"version": "v20.07.2",
"uptime": 2050,
"lastEcho": 1607358147
},
{
"instance": "zero",
"address": "10.30.0.180:5080",
"status": "healthy",
"group": "0",
"version": "v20.07.2",
"uptime": 5430,
"lastEcho": 1607358147
},
{
"instance": "zero",
"address": "10.30.0.180:5080",
"status": "healthy",
"group": "0",
"version": "v20.07.2",
"uptime": 5430,
"lastEcho": 1607358147
},
{
"instance": "alpha",
"address": "10.30.0.180:7080",
"status": "healthy",
"group": "1",
"version": "v20.07.2",
"uptime": 1797,
"lastEcho": 1607358147,
"ongoing": [
"opRollup"
],
"ee_features": [
"backup_restore"
]
}
]
- Why do I have 2 zeros with same IP here? 10.30.0.180:5080?
//10.30.0.26:7080
[
{
"instance": "zero",
"address": "10.30.0.26:5080",
"status": "healthy",
"group": "0",
"version": "v20.07.2",
"uptime": 3583,
"lastEcho": 1607358240
},
{
"instance": "zero",
"address": "10.30.0.180:5080",
"status": "healthy",
"group": "0",
"version": "v20.07.2",
"uptime": 5522,
"lastEcho": 1607358240
},
{
"instance": "zero",
"address": "10.30.0.191:5080",
"status": "healthy",
"group": "0",
"version": "v20.07.2",
"uptime": 6105,
"lastEcho": 1607358240
},
{
"instance": "alpha",
"address": "10.30.0.180:7080",
"status": "healthy",
"group": "1",
"version": "v20.07.2",
"uptime": 1888,
"lastEcho": 1607358240
},
{
"instance": "alpha",
"address": "10.30.0.26:7080",
"status": "healthy",
"group": "1",
"version": "v20.07.2",
"uptime": 2144,
"lastEcho": 1607358240,
"ongoing": [
"opRollup"
],
"ee_features": [
"backup_restore"
]
}
]
- Here 3 zeros with distinct IP’s here, seems ok, but, what about question 2?
That is the cluster when I navigated over ratel (I suppose it uses the same endpoints above, huh)
- How am I able to know which alpha should I point in order to run query against it? I Just realized if I connect via ratel in alpha 10.30.0.26:8080 (which is the leader) it returns me nothing, also, the Schema is missing too many predicates.
schema: 10.30.0.26:8080
schema: 10.30.0.180:8080
Also, there is a last thing that would block my usage of dgraph in a prod environment.
- After a bulk load of a 10GB rfd file, (around 47M items), whenever an alpha goes down, and a new one is started or becoming a leader, does it rebuild ALL the indexes? is it a expected behaviour? It is taking so long