Create gluster cluster API call doesn't work #400
Comments
|
@fbalak , Need to add |
|
@nthomas-redhat ok, thanks. I will later test it. |
|
I tried to create cluster but server returned: Response from Used curl:
Tested with: |
|
@r0h4n when the specified node (in this case ip_gl1, ip_gl2, ip_gl3) is not available under |
|
@fbalak can you check whether it there's "/indexes/ip/" in etcd for each of node (ip_gl1, ip_gl2, ip_gl3) |
|
There is no |
|
@fbalak Please attach a dump of your etcd data. This will need further inspection, also check if there's data in etcd "/networks/" |
|
There is no Dumps: Dumps from /nodes and /messages are too large. Maybe I can grant you an access to etcd, so you can look for yourself. |
|
When I tested it today it created job, but it failed. After this CC: @mkudlej Tested with: |
|
@fbalak In above comment, the last message says "Job failed Error setting up gluster node: Error setting up gluster node"
|
|
@fbalak Any updates? |
|
I have running Tested with: |
|
@fbalak For CreateCluster in case of Gluster you need to tag any 1 of the node-agents from the detected_cluster as "provisioner/gluster" (in /etc/tendrl/node-agent/node-agent.conf.yaml" Please verify this More details at Pt.4 : https://github.com/Tendrl/documentation/wiki/Tendrl-Package-Installation-Reference#storage-node-installation |
|
Yes, it is tagged. From |
|
@fbalak Please provide payload for Job |
|
|
|
@fbalak The "provisioner/gluster" tag is given to all the 3 nodes, only one node from the detected cluster has to be given that tag, can you retry after making changes and on a new etcd instance? |
|
Cluster creation still doesn't work for me. Services tendrl-node-agent and glusterd are running. curl command I used: On gluster node with provisioning appered new tag "provisioner/2c0f4770-2547-4ddf-a328-be8b96f2058f" as seen in output of NodeContext from etcd (other hosts don't have the tag): Tested with: |
|
@fbalak the new tag "provisioner/2c0f4770-2547-4ddf-a328-be8b96f2058f" indicates that python-gdeploy and gdeploy were installed on that node. This is good progress @nnDarshan @shtripat please debug the failures of above job, @fbalak feel free to schedule bluejeans or google hangout with them to debug the issue faster |
|
@r0h4n ok, thank you. |
|
@fbalak Tendrl tries to create password-less ssh from provisioner node to all the other nodes(including itself). |
|
@fbalak /v2/keys/messages/events in etcd should have more info about the error. |
|
It seems working now. --> closing |
When I tried creating cluster via API as documented I received
job_idbut the job is never finished. It's state remains asnew. I can't access job's messages either.Used curl:
curl -X POST -H 'Authorization: Bearer c06b8ede913fde12013828d5210461345b8830d8108c9b47727509397dea4a96' -d '{"sds_name":"gluster","sds_version":"3.9.1", "sds_parameters":{"name":"MyCluster","cluster_id":"mojeclusterid", "public_network":"server_ip/22","cluster_network":"server_ip/22"}, "node_identifier":"ip","node_configuration":{"gl1_ip":{"role": "glusterfs/node", "provisioning_ip":"gl1_ip"},"gl2_ip":{"role":"glusterfs/node", "provisioning_ip":"gl2_ip"},"gl3_ip":{"role":"glusterfs/node", "provisioning_ip":"gl3_ip"}}}' http://server_ip/api/1.0/CreateClusterFrom etcd:
{"action":"get","node":{"key":"/queue/752da924-31b2-40dc-bb30-678375aa7629","dir":true,"nodes":[{"key":"/queue/752da924-31b2-40dc-bb30-678375aa7629/status","value":"new","modifiedIndex":43283,"createdIndex":43283},{"key":"/queue/752da924-31b2-40dc-bb30-678375aa7629/payload","value":"{\"job_id\":\"752da924-31b2-40dc-bb30-678375aa7629\",\"status\":\"new\",\"name\":\"CreateCluster\",\"run\":\"tendrl.flows.CreateCluster\",\"type\":\"node\",\"created_from\":\"API\",\"created_at\":\"2017-04-26T12:12:21Z\",\"username\":\"admin\",\"parameters\":{\"TendrlContext.sds_name\":\"gluster\",\"TendrlContext.sds_version\":\"3.9.1\",\"TendrlContext.cluster_name\":\"MyCluster\",\"TendrlContext.cluster_id\":\"mojeclusterid\",\"Node[]\":[\"f83a3427-4e49-47a6-85e6-48f5d9fe8622\",\"8e6f4d79-f7dd-4666-865e-a0e53d4356c2\",\"f73d0f8b-cbcd-432f-9a48-7709f8515c4e\"],\"Cluster.public_network\":\"server_ip/22\",\"Cluster.cluster_network\":\"gl1_ip/22\",\"Cluster.node_configuration\":{\"f83a3427-4e49-47a6-85e6-48f5d9fe8622\":{\"role\":\"glusterfs/node\",\"provisioning_ip\":\"gl1_ip\"},\"8e6f4d79-f7dd-4666-865e-a0e53d4356c2\":{\"role\":\"glusterfs/node\",\"provisioning_ip\":\"gl2_ip\"},\"f73d0f8b-cbcd-432f-9a48-7709f8515c4e\":{\"role\":\"glusterfs/node\",\"provisioning_ip\":\"gl3_ip\"}},\"TendrlContext.integration_id\":\"bcb06eff-bbe8-438c-9dc5-848392726205\"},\"tags\":[\"provisioner/gluster\"]}","modifiedIndex":43284,"createdIndex":43284}],"modifiedIndex":43283,"createdIndex":43283}}Where server_ip, gl1_ip, gl2_ip and gl3_ip stand for valid machine ip addresses. On gluster machines is installed gluster but they are not in a trusted pool. Tendrl recognizes each of those machines as a single cluster (3 clusters).
Tested with:
tendrl-commons-1.2.3-04_26_2017_14_09_12.noarch
tendrl-dashboard-1.2.3-04_26_2017_04_11_04.noarch
tendrl-api-1.2.3-04_26_2017_01_51_04.noarch
tendrl-node-agent-1.2.3-04_26_2017_00_01_05.noarch
tendrl-node-monitoring-1.2.3-04_26_2017_04_25_04.noarch
The text was updated successfully, but these errors were encountered: