Trouble deploying 1.0.2

We were using the earlier version based on ES6 on top of kubernetes with helm. We are struggling to adapt our helm chart to deploy the ES7 based 1.0.2. We are using the following config on our masters:

cluster.name: "opendistro-es-qa"
cluster.initial_master_nodes:       
  - opendistro-es-qa-master-0
  - opendistro-es-qa-master-1
  - opendistro-es-qa-master-2
network.host: "0.0.0.0"
discovery.seed_hosts: "opendistro-es-qa-master-headless"
cluster.routing.allocation.disk.threshold_enabled: true
node.max_local_storage_nodes: 1
node.ingest: false
node.data: false
node.master: true
opendistro_security.allow_unsafe_democertificates: false
opendistro_security.allow_default_init_securityindex: true
opendistro_security.audit.type: internal_elasticsearch
opendistro_security.enable_snapshot_restore_privilege: true
opendistro_security.check_snapshot_restore_write_privileges: true
opendistro_security.restapi.roles_enabled: ["all_access", "security_rest_api_access"]
opendistro_security.ssl.transport.enforce_hostname_verification: false
opendistro_security.ssl.http.enabled: true
opendistro_security.ssl.transport.pemcert_filepath: elk-transport-crt.pem
opendistro_security.ssl.transport.pemkey_filepath: elk-transport-key.pem
opendistro_security.ssl.transport.pemtrustedcas_filepath: elk-transport-root-ca.pem
opendistro_security.ssl.http.pemcert_filepath: elk-rest-crt.pem
opendistro_security.ssl.http.pemkey_filepath: elk-rest-key.pem
opendistro_security.ssl.http.pemtrustedcas_filepath: elk-rest-root-ca.pem
opendistro_security.authcz.admin_dn: [CN=Service Bind OpenDistro,OU=ServiceAccounts,OU=Core,DC=Test,DC=Lab]
# testing.
opendistro_security.nodes_dn: [/CN=.*/]

There are no config options specified in the env vars.

Yet when the master comes up, I get this error.

org.elasticsearch.bootstrap.StartupException: java.lang.IllegalArgumentException: it is forbidden to set both [discovery.seed_hosts] and [discovery.zen.ping.unicast.hosts]

There was an environment variable setting after all

I have also been facing similar issue on my client node while deploying ODFE using helm chart.
I have 1 master, 1 data and 1 client node. While the master and data nodes are working as expected but the client node is coming up. The readiness probe is failing and upon checking logs i am seeing below error:
[2019-09-19T16:43:01,140][ERROR][c.a.o.s.a.BackendRegistry] [es-open-distro-helm-client-77bbcc8c7d-qvk55] Not yet initialized (you may need to run securityadmin)

I ran securityadmin script and i executed successfully but the client node is still not coming up.

1 Like