s9s-node

View and handle nodes.

Usage

s9s node {command} {options}

Command

Options

Graph Options

When providing a valid graph name together with the --stat option a graph will be printed with statistical data. Currently, the following graphs are available:

Node List

Using the --list and --long command-line options a detailed list of the nodes can be printed. Here is an example of such a list:

$ s9s node --list --long '192.168.1.1*'
STAT VERSION CID CLUSTER HOST PORT COMMENT
poM- 9.6.2 1 ft_postgresql_11794 192.168.1.117 8089 Up and running
coC- 1.4.2 1 ft_postgresql_11794 192.168.1.127 9555 Up and running
Total: 3

The list contains the following fields:

Node Format

The string that controls the format of the printed information about the nodes. When this command-line option is used the specified information will be printed instead of the default columns. The format string uses the % character to mark variable fields and flag characters as they are specified in the standard printf() C library functions. The % specifiers are ended by field name letters to refer to various properties of the nodes.

The %+12i format string for example has the +12 flag characters in it with the standard meaning the field will be 12 character-wide and the + or - sign will always be printed with the number. The properties of the node are encoded by letters. In the %16D for example, the letter D encodes the “data directory” field, so the full path of the data directory on the node will be substituted. Standard \ notation is also available, \n for example encodes a new-line character.

The s9s-tools support the following fields:

Examples

List all nodes:

$ s9s node --list --long
ST  VERSION                  CID CLUSTER        HOST       PORT COMMENT
go- 10.1.22-MariaDB-1~xenial   1 MariaDB Galera 10.0.0.185 3306 Up and running
co- 1.4.1.1856                 1 MariaDB Galera 10.0.0.205 9500 Up and running
go- 10.1.22-MariaDB-1~xenial   1 MariaDB Galera 10.0.0.209 3306 Up and running
go- 10.1.22-MariaDB-1~xenial   1 MariaDB Galera 10.0.0.82  3306 Up and running
Total: 4

Print the configuration for a node:

$ s9s node --list-config --nodes=10.0.0.3
...
mysqldump   max_allowed_packet                     512M
mysqldump   user                                   backupuser
mysqldump   password                               nWC6NSm7PnnF8zQ9
xtrabackup  user                                   backupuser
xtrabackup  password                               nWC6NSm7PnnF8zQ9
MYSQLD_SAFE pid-file                               /var/lib/mysql/mysql.pid
MYSQLD_SAFE basedir                                /usr/
Total: 71

The following example shows how a node in a given cluster can be restarted. When this command executed a new job will be created to restart a node. The command-line tool will stop and show the job messages until the job is finished:

$ s9s node \
        --restart \
        --cluster-id=1 \
        --nodes=192.168.1.117 \
        --log

Change a configuration value for a PostgreSQL server:

$ s9s node \
        --change-config \
        --nodes=192.168.1.115 \
        --opt-name=log_line_prefix \
        --opt-value='%m '

Push a configuration option inside my.cnf (max_connections=500) on node 10.0.0.3:

$ s9s node \
        --change-config \
        --nodes=10.0.0.3 \
        --opt-group=mysqld \
        --opt-name=max_connections \
        --opt-value=500

Import two existing Keepalived nodes (192.168.20.56 and 192.168.20.57) where the virtual IP is 192.168.20.59 into cluster ID 13:

# import primary keepalived
$ s9s node \
        --register \
        --nodes="keepalived://192.168.20.56" \
        --virtual-ip=192.168.20.59 \
        --cluster-id=13 \
        --wait

# import secondary keepalived
$ s9s node \
        --register \
        --nodes="keepalived://192.168.20.57" \
        --virtual-ip=192.168.20.59 \
        --cluster-id=13 \
        --wait

Listing the Galera hosts. This can be done by filtering the list of nodes by their properties:

$ s9s node \
        --list \
        --long \
        --properties="class_name=CmonGaleraHost"

Create a set of graphs, one for each node shown in the terminal about the load on the hosts. If the terminal is wide enough the graphs will be shown side by side for a compact view:

$ s9s node \
        --stat \
        --cluster-id=1 \
        --begin="08:00" \
        --end="14:00" \
        --graph=load

The density function can also be printed to show what were the typical values for the given statistical data. The following example shows what was the typical values for the user mode CPU usage percent:

$ s9s node \
        --stat \
        --cluster-id=2 \
        --begin=00:00 \
        --end=16:00 \
        --density \
        --graph=cpuuser

The following example shows how a custom list can be created to show some information about the CPU(s) in some specific hosts:

$ s9s node \
        --list \
        --node-format="%N %U CPU %c Cores %6.2u%% %Z\n" 192.168.1.191 192.168.1.195
192.168.1.191 2 CPU 16 Cores 22.54% Intel(R) Xeon(R) CPU L5520 @ 2.27GHz
192.168.1.195 2 CPU 16 Cores 23.12% Intel(R) Xeon(R) CPU L5520 @ 2.27GHz

The following list shows some information about the memory, the total memory and the memory available for the applications to allocate (including cache and buffer with the free memory):

$ s9s node \
        --list \
        --node-format="%4.2m GBytes %4.2fm GBytes %N\n"
16.00 GBytes 15.53 GBytes 192.168.1.191
47.16 GBytes 38.83 GBytes 192.168.1.127

Set a node to read-write mode if it was previously set to read-only mode:

$ s9s node \
        --set-read-write \
        --cluster-id=1 \
        --nodes=192.168.0.78 \
        --log

Copy configuration file(s) from a PostgreSQL server 192.168.0.232 into the localhost:

$ s9s node \
        --pull-config \
        --nodes="192.168.0.232" \
        --output-dir="tmp"