I have deployed 3 vROPS 6.2 OVA.For the master I have configured running
curl -s -v -w %{http_code} -i -k -H Content-Type:application/json -d@/tmp/vrops-config-2449/role-body -u admin:VMware123* -b /tmp/vrops-config-2449/cookie -X POST https://10.1.1.64/casa/deployment/slice/role
With this autogenerated role-body json file,
[
{
"slice_address": "10.1.1.64",
"admin_slice": "10.1.1.64",
"user_id": "admin",
"password":"VMware123*",
"slice_roles": ["ADMIN","DATA","UI"]
}
]
And it worked fine.
My problems appeared when I tried to configure the datanode and the remote collector. I used the same
curl -s -v -w %{http_code} -i -k -H Content-Type:application/json -d@/tmp/vrops-config-2449/role-body -u admin:VMware123* -b /tmp/vrops-config-2449/cookie -X POST https://10.1.1.64/casa/deployment/slice/role
With this role-body json file, for Remote collector ,
[
{
"slice_address": "10.1.1.226",
"admin_slice": "10.1.1.64",
"user_id": "admin",
"password":"VMware123*",
"slice_roles": ["REMOTE_COLLECTOR"]
}
]
and this role-body json file-for datanode,
[
{
"slice_address": "10.1.1.64",
"admin_slice": "10.1.1.64",
"user_id": "admin",
"password":"VMware123*",
"slice_roles": ["DATA","UI"]
}
]
this is the output:
Configuring vROps DataNode Role ...
curl -s -v -w %{http_code} -i -k -H Content-Type:application/json -d@/tmp/vrops-config-2449/role-body -u admin:VMware123* -b /tmp/vrops-config-2449/cookie -X POST https://10.1.1.64/casa/deployment/slice/role
* About to connect() to 10.1.1.64 port 443 (#0)
* Trying 10.1.1.64... connected
* Connected to 10.1.1.64 (10.1.1.64) port 443 (#0)
* Initializing NSS with certpath: /etc/pki/nssdb
* warning: ignoring value of ssl.verifyhost
* skipping SSL peer certificate verification
* SSL connection using TLS_RSA_WITH_AES_256_CBC_SHA
* Server certificate:………
* Server auth using Basic with user 'admin'
> POST /casa/deployment/slice/role HTTP/1.1
> Authorization: Basic YWRtaW46Vk13YXJlMTIzKg==
> User-Agent: curl/7.19.7 (x86_64-unknown-linux-gnu) libcurl/7.19.7 NSS/3.12.7.0 zlib/1.2.3 libidn/1.18 libssh2/1.2.2
> Host: 10.1.1.64
> Accept: */*
> Content-Type:application/json
> Content-Length: 181
>
< HTTP/1.1 500 Internal Server Error
< Connection: close
< Content-Type: application/json;charset=ISO-8859-1
{"error_message_key":"general.failure","error_arguments":["com.vmware.vcops.casa.exception.CasaException: Error validating credentials"]}500
NOTE: The credentials for the vROPs Cluster/Master were correct introduced in cURL and json.
After seeing the error message ,I have decided to modify the json file, for testing purpose and the error message was something I expecting.
Next I decided to modify the json file and run again the same link:
With this autogenerated role-body json file:
[
{
"slice_address": "10.1.1.226",
"admin_slice": "10.1.1.64",
"slice_roles": ["REMOTE_COLLECTOR"]
}
]
Configuring vROps DataNode Role ...
curl -s -v -w %{http_code} -i -k -H Content-Type:application/json -d@/tmp/vrops-config-2521/role-body -u admin:VMware123* -b /tmp/vrops-config-2521/cookie -X POST https://10.1.1.64/casa/deployment/slice/role
< HTTP/1.1 500 Internal Server Error>
{"error_message_key":"general.failure","error_arguments":["java.lang.IllegalArgumentException: Either username/password, or shared passphrase must be provided."]}500
My two cents on my tests with vROPS 6.2:
1- First of all, if I add a datanode using the Webinterface Wizzard, I can expand successfully the Cluster with the datanode.
Here comes the weird thing: if I want to change the role configuration (from datanode to remote collector), for testing only, of the datanode mentioned above, using the same CURL: curl -s -v -w %{http_code} -i -k -H Content-Type:application/json -d@/tmp/vrops-config-2521/role-body -u admin:VMware123* -b /tmp/vrops-config-2521/cookie -X POST https://10.1.1.64/casa/deployment/slice/role , I can successfully do it.
NO "Error validating credentials" is returned. It just switches the datanote into a remote collector and vice-versa. The confirmation of this role change is in the webinterface/admin as well as when I run service vmware-vcops status.
2-After debugging and testing I have a couple of questions:
A- VMware might have changed the json structure, the json (role_body) expected by the master? Eg: "user_id" could might have been changed to
"userName" or anything else?
B- although I am using curl -k for skipping SSL peer certificate verification, is master actually trusts the new Slice(datanode or remote collector)?
Can some one explain this behaviour?
Thanks.
