cbopinfo

cbopinfo is a command-line tool that is provided in the Couchbase Autonomous Operator package. It is used to collect resources and logs via the Kubernetes API in order to support the Operator and Couchbase Server.

By default, cbopinfo will collect resources for all Couchbase clusters defined in the selected namespace. Certain resource types such as services, pods, and persistent volume claims are scoped to those created by the Operator. The scope can be further restricted to only the clusters specified on the command line.

Other unscoped resources are collected without any filtering, such as roles, role bindings, and service accounts, as they cannot be distinguished as belonging to a particular Couchbase cluster.

Any secrets that are collected have their data redacted so only the presence of a secret or a key within it are visible.

The output is a compressed archive of resource YAML manifests organized by namespace, resource type, and name. Logs and events are also collected and collated with resource configuration where relevant.

The cbopinfo command can also collect and download Couchbase server logs from running pods and orphaned persistent log volumes.

The cbopinfo command requires permission to read resources related to the cluster in order to function correctly e.g. deployments, pods, Couchbase clusters, etc. When running this command ensure it is run by a user with sufficient permission to collect the required resources. If permission is denied to access a particular resource type the command will notify the user and continue with the collection on a best attempt basis. In general the required permissions are similar to those required by the operator but read-only.

Installation

Make sure that you have downloaded the Operator package and unpacked it on the same computer where you normally run kubectl or oc.

After you unpack the download, the resulting directory will be titled something like couchbase-autonomous-operator-kubernetes_x.x.x-linux_x86_64.

macOS
Linux
Windows

Open a Terminal window and go to the directory where the cbopinfo binary is located:
```
$ cd couchbase-autonomous-operator-kubernetes_x.x.x-macos_x86_64/bin/
```
Make the cbopinfo binary executable:
```
$ chmod +x ./cbopinfo
```

Move the binary into your PATH:

$ sudo mv ./cbopinfo /usr/local/bin/cbopinfo

Open a command prompt and go to the directory where the cbopinfo binary is located:
```
$ cd couchbase-autonomous-operator-kubernetes_x.x.x-linux_x86_64/bin/
```
Make the cbopinfo binary executable:
```
$ chmod +x ./cbopinfo
```

Move the binary into your PATH:

$ sudo mv ./cbopinfo /usr/local/bin/cbopinfo

Open a command prompt and go to the directory where the cbopinfo binary is located:
```
$ cd couchbase-autonomous-operator-kubernetes_x.x.x-windows_x86_64\bin\
```
Add the cbopinfo binary into your PATH.

The cbopinfo command conforms to Kubernetes cli-runtime specification and may be renamed to kubernetes-couchbase-logs and installed in your $PATH. It can then be invoked as a kubectl plugin with kubectl couchbase logs. This tightly integrates it with your normal workflow.

Options

--all

If specified, this flag ignores any scoping rules for a particular resource type, and collects all instances within the namespace.

--collectinfo

If specified, this flag will trigger collection of Couchbase logs via the cbcollect_info command locally on each pod belonging to a Couchbase cluster. By default an interactive menu allows selection of logs to collect and also allows the user to specify whether the logs should be redacted of sensitive information. The logs will be downloaded to the current working directory once collection is complete.

For log volumes that were previously attached to ephemeral pods, cbopinfo will create a temporary pod (as defined by the --server-image option), run cbcollect_info and download the logs before finally deleting the temporary pod.

Example Output of --collectinfo

$ ./build/bin/cbopinfo --collectinfo
Detected resources for log collection:
┌────┬─────────────────┬──────┬──────────┐
│ ID │ Pod             │ Type │ Detached │
├────┼─────────────────┼──────┼──────────┤
│ 0  │ cb-example-0001 │ pod  │          │
│ 1  │ cb-example-0003 │ pod  │          │
│ 2  │ cb-example-0004 │ pod  │          │
└────┴─────────────────┴──────┴──────────┘
Please select volumes to collect [e.g. 1,2,5-6 or leave blank for all]: 1-2
Please select whether to enable log redaction [Y/n]:
Server logs downloaded to the following files:
    cbinfo-default-cb-example-0002-20181016T131730+0100-redacted.zip
    cbinfo-default-cb-example-0001-20181016T131730+0100-redacted.zip
Wrote cluster information to cbopinfo-20181016T131730+0100.tar.gz

--collectinfo-list

If specified with the --collectinfo option, this will write out the list of available collections to the console and exit. It is designed to be used with automated tooling to poll for and select logs for later collection with the --collectinfo-collect option.

The format of the --collectinfo-list option is as follows:

{
    "items": [
        {
            "name": "cb-example-0001",
            "type": "pod"
        },
        {
            "name": "cb-example-0003",
            "type": "pod"
        },
        {
            "name": "cb-example-0004",
            "type": "pod"
        },
        {
            "detached": "2018-10-16T11:27:45Z",
            "name": "cb-example-0000",
            "type": "persistentVolumeClaim"
        },
        {
            "detached": "2018-10-16T10:58:34Z",
            "name": "cb-example-0002",
            "type": "persistentVolumeClaim"
        }
    ]
}

The name is the name of the Pod that either can be collected from, or the Pod that a log volume was attached to. The type specifies whether the logs are from a running pod or a log persistentVolumeClaim. The detached field is only specified for collections of type persistentVolumeClaim and records when the log volume was detected as orphaned from its Pod.

--collectinfo-collect <collection>

If specified with the --collectinfo option, this will define the set of Couchbase server logs to collect. This option will override and not display the interactive menu shown by default with the --collectinfo option.

The collection is specified as a comma separated list of log indices, indexed from zero. For example to collect the logs for cb-example-0004 in the example for the --collectinfo-list option this would be index 2. You may specify ranges of indices, for example 1-2. You may also specify all to collect all available logs. Valid collection specifications could be:

1             # Collect log index 1 only
1,2,4         # Collect log indices 1, 2 & 4
6-8           # Collect log indices 6, 7 & 8
1,3,6-8,12    # Collect log indices 1, 3, 6, 7, 8 & 12
all           # Collect all available logs

--collectinfo-redact

If specified with the --collectinfo option, this will indicate that only redacted logs should be collected and downloaded.

--operator-image <image>

If specified this sets the Operator image name. It is used to match containers running the Operator for collection of run time information. If not set it will default to couchbase/operator:2.1.0.

--operator-rest-port <port>

If specified this sets the Operator REST API port. It is used for collection of run time information. If not set it will default to 8080.

--server-image <image>

If specified this defines the Couchbase Server image to use when collecting logs from detached log volumes. If not set, it will default to couchbase/server:enterprise-6.6.0

--system

If specified, this flag will additionally collect all resources from the kube-system namespace, which may be useful in diagnosing problems with cluster-level roles, custom resource definitions, or system services.

INFO: cbopinfo supports all standard configuration options available in cli-runtime e.g. --namespace, --context, etc.

Examples

To perform a basic collection of all Couchbase clusters and their associated resources within a name space:

$ cbopinfo
Wrote cluster information to cbopinfo-20180629T123256+0100.tar.gz

To collect basic logs on OpenShift from a specific name space:

$ oc login -u developer
Authentication required for https://192.168.99.100:8443 (openshift)
Username: developer
Password:
Login successful.

You have access to the following projects and can switch between them with 'oc project ':

    myproject

Using project "myproject".

$ cbopinfo --namespace myproject --operator-image couchbase/operator:2.1.0-1
Wrote cluster information to cbopinfo-20180801T141838+0100.tar.gz

To collect all Couchbase clusters, their associated resources within a namespace, Couchbase server logs, and Kubernetes system logs:

$ cbopinfo --collectinfo --collectinfo-collect all --system
Server logs downloaded to the following files:
    cbinfo-default-cb-example-0002-20181016T145816+0100.zip
    cbinfo-default-cb-example-0000-20181016T145816+0100.zip
    cbinfo-default-cb-example-0004-20181016T145816+0100.zip
    cbinfo-default-cb-example-0003-20181016T145816+0100.zip
    cbinfo-default-cb-example-0001-20181016T145816+0100.zip
Wrote cluster information to cbopinfo-20181016T145816+0100.tar.gz

To interactively collect server logs:

$ ./build/bin/cbopinfo --collectinfo
Detected resources for log collection:
┌────┬─────────────────┬──────┬──────────┐
│ ID │ Pod             │ Type │ Detached │
├────┼─────────────────┼──────┼──────────┤
│ 0  │ cb-example-0001 │ pod  │          │
│ 1  │ cb-example-0003 │ pod  │          │
│ 2  │ cb-example-0004 │ pod  │          │
└────┴─────────────────┴──────┴──────────┘
Please select volumes to collect [e.g. 1,2,5-6 or leave blank for all]: 1-2
Please select whether to enable log redaction [Y/n]:
Server logs downloaded to the following files:
    cbinfo-default-cb-example-0002-20181016T131730+0100-redacted.zip
    cbinfo-default-cb-example-0001-20181016T131730+0100-redacted.zip
Wrote cluster information to cbopinfo-20181016T131730+0100.tar.gz