Interacting with Domino server process within Domino container
Important
This is only applicable when you install Volt MX Go to a development or test only environment.
The Volt MX Go installation in a development or test only environment offers a Helm chart to run a Domino server enabled with the Domino REST API (DRAPI) in a Kubernetes environment. This topic offers guidance and tips on working with the Domino container.
Container basics
When you deploy the Volt MX Go Domino REST API Helm chart, you instruct Kubernetes to create a pod. The Helm chart provides the specifics for what program (Image) will run when the pod starts, the details around what ports to expose, such as port 80 for Domino HTTP server, port 8880 for Domino REST API, and the memory and CPU allocations among other things.
The chart also tells Kubernetes about liveness and readiness probes. Kubernetes runs these probes periodically to determine container health. If a readiness probe fails more than a specified threshold, the container is marked as unready and won't receive any HTTP traffic. If a liveness probe fails many times, Kubernetes assumes the container has unrecoverable errors and then kills and restarts the container.
These probes are important to understand as some operations require you to stop and restart Domino. If Domino fails to answer liveness probes more than the specified threshold, such as when Domino is stopped, Kubernetes may remove the container and start a new one. By default, this period is 10 minutes for the Domino container.
The Domino data directory resides in /local/notesdata within the Domino container. This directory is actually a mount point for a persistent, initially empty file system created by Kubernetes.
Tip
For more information, see persistent volumes.
The directory is used to store the Domino server configuration, Notes databases, templates, and other details specific to the Domino server instance. During initial startup, Domino sees the empty directory and triggers the Domino One Touch setup process to populate the directory with databases, templates, and configuration files. Logs are in /local/notesdata/IBM_TECHNICAL_SUPPORT/.
Domino runs under the Unix user ID of hcl with primary group of root.
Pod vs Container
A pod and a container are related concepts, but serve different purposes and have distinct characteristics.
Container:
- Lightweight, standalone, and executable software package that includes everything needed to run a piece of software, including the code, runtime, system tools, and libraries.
- Provide a consistent and isolated environment for running applications. They're designed to be portable and can run consistently across different environments, such as development, testing, and production.
- Typically, created using containerization technologies like Docker, and they encapsulate individual applications or services.
Pod:
- Smallest deployable unit in Kubernetes, which is a container orchestration platform.
-
Contains one or more containers that are coupled and share the same network namespace, IPC namespace, and storage volumes.
Note
Containers within the same pod share the same resources and can communicate with each other using localhost, making them suitable for co-locating integrated services.
-
Often used to group containers that need to work together, such as a web server and a sidecar for logging.
Key differences:
- A container is a single lightweight unit that encapsulates an application and its dependencies.
- A pod is a higher-level concept in Kubernetes and can contain one or more coupled containers scheduled together on the same host.
- Containers package and run individual applications or services, while pods group containers that need to work together and share certain resources.
Kubernetes uses pods as its fundamental scheduling unit because they provide a way to manage co-located containers and offer flexibility for deploying complex applications with multiple services.
Pod names
Pods have names that usually consist of developer defined prefix and runtime defined unique suffix. Together, they make each pod name unique. Because of this, you often have to get a list of the running pods to determine the exact name before taking action on the pod. Use the kubectl command get pods as shown here:
kubectl get pods -n mxgo
You should see output similar to this:
NAME READY STATUS RESTARTS AGE
drapi-6949c45b8-wghbz 3/3 Running 0 8d
mysql-0 1/1 Running 0 8d
foundry-db-update-zn5vg 0/1 Completed 0 7d23h
voltmx-foundry-apiportal-c874cb6d-l2k8h 1/1 Running 0 7d22h
voltmx-foundry-integration-76dc89b57-qmq2t 1/1 Running 0 7d22h
voltmx-foundry-identity-85f9854f8b-n6b2d 1/1 Running 0 7d22h
voltmx-foundry-console-664c75c6b5-8lm6v 1/1 Running 0 7d22h
Note
If the pod is restarted for any reason, the pod will have a new name, so you need to get the list of running pods to know the exact name.
Copy files into and out of Domino container
Using the kubectl cp command, you can copy files into and out of the container. For example, if you want to copy the domino administrative ID out of the Domino data directory, you use a command like this:
kubectl cp domino-drapi-6fc766ccd4-55d9n:/local/notesdata/admin.id domino-id.id -n mxgo -c drapi
Here, the cp command copies the file /local/notesdata/admin.id out of the container domino-drapi-6fc766ccd4-55d9n to the file domino-id.id in the current working directory. The flag -n mxgo indicates that the Kubernetes namespace is mxgo and the flag -c drapi indicates using the drapi container within the Domino DRAPI pod.
Similarly, you could copy the local file foo.bar into the container's /tmp directory like this:
kubectl cp foo.bar domino-drapi-6fc766ccd4-55d9n:/tmp/foo.bar -n mxgo -c drapi
Refer to Kubernetes documentation for more details on cp.
Interactive command line access with Domino
You can get a bash prompt within the Domino container much like when you SSH into a remote Linux host. You do this with the kubectl exec command. You may start a bash command session like this:
kubectl exec -it domino-drapi-6fc766ccd4-55d9n -n mxgo -c drapi -- bash
The response should look like the following, which indicates you are now in a bash shell command prompt within the Domino REST API container:
[notes@domino-drapi-6fc766ccd4-55d9n /]$
You can use the nashcom Domino Start Script to interact with the Domino server. This script is already installed and available for use.
You can show the status of the Domino server by using this command:
domino status
You should see:
Domino Server is running (notes)
You can attach to the Domino console with the live command and see something like this:
[notes@domino-drapi-6fc766ccd4-55d9n /]$ domino live
Using Domino config File [/etc/sysconfig/rc_domino_config]
--- Live Console for notes ---
To close console, always type 'close' or 'stop'.
[000582:000002-00007FBF5156C740] 10/09/2023 04:39:48 PM Directory Catalog entitlements.nsf needs rebuilding (Replication to all mobile clients will take longer): Rebuilding due to no replication history
[006013:000002-00007FA7624F1740] 10/09/2023 04:39:48 PM Chronos: Performing hourly full text indexing
[006013:000002-00007FA7624F1740] 10/09/2023 04:39:55 PM Chronos: Full text indexer terminating
[000588:000002-00007FF2B92FC740] 10/09/2023 04:45:42 PM Admin Process: Searching Administration Requests database
[009792:000002-00007F0024E1A740] 10/09/2023 05:39:53 PM Chronos: Performing hourly full text indexing
[009792:000002-00007F0024E1A740] 10/09/2023 05:39:59 PM Chronos: Full text indexer terminating
[000588:000002-00007FF2B92FC740] 10/09/2023 05:45:41 PM Admin Process: Searching Administration Requests database
[013563:000002-00007FF34BA4B740] 10/09/2023 06:39:53 PM Chronos: Performing hourly full text indexing
[013563:000002-00007FF34BA4B740] 10/09/2023 06:39:59 PM Chronos: Full text indexer terminating
[000588:000002-00007FF2B92FC740] 10/09/2023 06:45:42 PM Admin Process: Searching Administration Requests database
Once in the console, you can run other Domino console commands such as show tasks:
show tasks
[000191:000015-00007FEDB3E7C700] Task Description
[000191:000015-00007FEDB3E7C700] Database Server Perform console commands
[000191:000015-00007FEDB3E7C700] Database Server Listen for connect requests on TCPIP
[000191:000015-00007FEDB3E7C700] Database Server Load Monitor is idle
[000191:000015-00007FEDB3E7C700] Database Server Database Directory Manager Cache Refresher is idle
[000191:000015-00007FEDB3E7C700] Database Server Organization Name Cache Refresher is idle
[000191:000015-00007FEDB3E7C700] Database Server Log Purge Task is idle
[000191:000015-00007FEDB3E7C700] Database Server Idle task
[000191:000015-00007FEDB3E7C700] Database Server Idle task
[000191:000015-00007FEDB3E7C700] Database Server log.nsf Commit Thread
[000191:000015-00007FEDB3E7C700] Database Server Perform Database Cache maintenance
[000191:000015-00007FEDB3E7C700] Database Server Idle task
[000191:000015-00007FEDB3E7C700] Database Server Platform Stats is idle
[000191:000015-00007FEDB3E7C700] Database Server Metric Task
[000191:000015-00007FEDB3E7C700] Database Server Shutdown Monitor
[000191:000015-00007FEDB3E7C700] Database Server Process Monitor
[000191:000015-00007FEDB3E7C700] LDAP Server Listen for connect requests on TCP Port:389
[000191:000015-00007FEDB3E7C700] LDAP Server Utility task
[000191:000015-00007FEDB3E7C700] Router Utility: Idle
[000191:000015-00007FEDB3E7C700] Router MailEvent: Idle
[000191:000015-00007FEDB3E7C700] Router Dispatch: Idle
[000191:000015-00007FEDB3E7C700] Router Dispatch: Idle
[000191:000015-00007FEDB3E7C700] Agent Manager Executive '1': Idle
Experienced Domino administrators know that sometimes you must run a process called fixup to fix a corrupt database. You can do this from here using the following command:
Load fixup <databasepath> <options>
You may disconnect from the console, and leave the Domino Server running, using the command close.
For the full set of server commands and syntax, see the HCL Domino Documentation.
Stop and start Domino server process
If you need to stop and restart the Domino server without killing the entire pod, you may exec into the Domino pod and issue the command to stop and then restart Domino. Here is an example with a typical output:
kubectl exec -it domino-drapi-6fc766ccd4-55d9n -n voltmxgo -c drapi -- bash
[notes@domino-drapi-6fc766ccd4-55d9n /]$ export PATH=$PATH:/opt/nashcom/startscript/
[notes@domino-drapi-6fc766ccd4-55d9n /]$ domino stop
Using Domino config File [/etc/sysconfig/rc_domino_config]
Stopping Domino for xLinux (notes)
KEEP started in task mode, no pre_shutdown action required
... waiting for shutdown to complete
... waiting 10 seconds
... waiting 20 seconds
KEEP started in task mode, no post_shutdown action required
Domino stopped (26 sec)
Domino for xLinux (notes) shutdown completed
Restart using the start command:
[notes@domino-drapi-6fc766ccd4-55d9n /]$ domino start
Using Domino config File [/etc/sysconfig/rc_domino_config]
Archived log file to '/local/notesdata/notes_231009_193905.log'
Removed LoadMon-Data '/local/notesdata/loadmon.ncf'
Starting Domino for xLinux (notes)
done PID is 20210
KEEP starting in task mode, no post_startup action required
[notes@domino-drapi-6fc766ccd4-55d9n /]$
You can also use the restart command.
Important
If you leave Domino stopped for too long, usually longer than 10 minutes, Kubernetes liveness tests detect this and may delete the pod and restart a new one.