Opennebula 5.0 cluster

Community Support
1

Hi ,
I have an issue while creating vm,

No system datastore meets capacity and SCHED_DS_REQUIREMENTS: “CLUSTERS/ID” = 100 & ( CLUSTER_ID=“100” )

I have created two clusters node1 and node2.
i use different vnet, host and system datastore for each cluster,
but use same cephdatastore for both cluster

onecluster show node1
CLUSTER 100 INFORMATION
ID : 100
NAME : node1

CLUSTER TEMPLATE
RESERVED_CPU="0.5"
RESERVED_MEM=“0.5”

HOSTS
3

VNETS
0

DATASTORES
0
104 cephds

onecluster show node2
CLUSTER 101 INFORMATION
ID : 101
NAME : node2

CLUSTER TEMPLATE
RESERVED_CPU="“
RESERVED_MEM=”"

HOSTS
5

VNETS
4

DATASTORES
104 cephds
107 system_ds

In my template i use cluster as below

SCHED_DS_REQUIREMENTS = "CLUSTER_ID="100""
SCHED_REQUIREMENTS = "CLUSTER_ID=“100"”

( or )

SCHED_DS_REQUIREMENTS = "CLUSTER_ID="101""
SCHED_REQUIREMENTS = "CLUSTER_ID=“101"”

Thanks in advance

2

Hi,

CLUSTER_ID will not work inside SCHED_DS_REQUIREMENTS. Did you copy that from our documentation? If so let me know and I’ll fix it.

In any case, the requirements will be set automatically. If the VM uses the vnet from one of the clusters, it will be deployed there because of the automatic SCHED_*REQUIREMENTS. In case of ambiguity, setting the SCHED_REQUIREMENTS (for hosts) will force the VM to be deployed in a specific cluster. In here you can use CLUSTER_ID.

3

Hi cmartin,

you are right SCHED_DS_REQUIREMENTS is not referred in documents. I met an issue when i use two clusters in Opennebula 4.14.2

No system datastore meets capacity and SCHED_DS_REQUIREMENTS: “CLUSTER_ID” = 100 & ( CLUSTER_ID=“100” ) & “CLUSTER_ID= 101” because of the automatic SCHED_*REQUIREMENTS

thats why i tried like that on Opennebula 4.14.2. so manually added the field.

i didn’t met any issue because of this field (SCHED_DS_REQUIREMENTS = "CLUSTER_ID=“100"”) on 4.14.2

in 5.0 after removed SCHED_DS_REQUIREMENTS field it works well
Thank you so much,