Node diagnostic command reference

This page lists the read-only commands available on an SC//Reliant™ platform node for inspecting its network and storage configuration, and describes what each one reports. It is a reference rather than a procedure. For the tasks these commands support, see Related documentation.

Every command on this page reads state and changes nothing. Making changes on a node is done through SC//Reliant Platform Manager and the automation framework, not by editing configuration directly.

Running these commands requires shell access to the node, with sudo where indicated. For the account settings and the SSH key this depends on, see Configuring shell access. For the SSH command itself and which node to run a given command from, see Connecting to a node.

Network commands

Command	Reports
`ip a`	Every network interface on the node, its state, and the addresses assigned to it.
`ip r`	The routing table, including the default gateway.
`bridge link`	Bridge membership and the VLANs carried on each bridge port.
`ss -tlnp`	Listening TCP sockets, with the process bound to each one.
`ping -c4 <host>`	Whether a host responds, and the round-trip time. The `-c4` limits it to four packets.

ip a

Lists interfaces with their state and addresses. Physical interfaces, bridges, and the virtual interfaces belonging to running virtual machines all appear here.

ip a

An interface carrying traffic reads state UP. An interface that exists but has no carrier reads state DOWN, which on an unused port is normal.

ip r

Prints the routing table. The line beginning default via identifies the gateway the node uses for traffic with no more specific route.

ip r

bridge link

Prints the ports attached to each bridge. On a node running virtual machines, each VM's virtual interface appears as a port on the bridge that carries its VLAN.

bridge link

ss -tlnp

Lists listening TCP sockets. The flags are -t for TCP, -l for listening only, -n for numeric ports rather than service names, and -p for the owning process. The process column requires sudo.

sudo ss -tlnp

ping

Sends four ICMP echo requests to a host and reports replies and round-trip times.

ping -c4 <host>

NOTE: A host that does not reply is not necessarily unreachable. Many devices and firewalls drop ICMP while still serving traffic normally, so treat a failed ping as one data point rather than proof.

Storage commands

Command	Reports
`lsblk`	Block devices in a tree, showing partitions, logical volumes, and mount points.
`sudo lvs`	Logical volumes, their volume group, and their size.
`sudo vgs`	Volume groups, their total size, and their free space.
`rvc sys storage`	Mounted filesystems with their device, type, total size, and percentage used.

lsblk

Prints block devices as a tree, so a physical disk, its partitions, and the logical volumes built on them appear in their hierarchy.

lsblk

lvs and vgs

lvs lists logical volumes and vgs lists the volume groups containing them. On a cluster node the logical volumes underneath DRBD devices appear here.

sudo lvs

sudo vgs

rvc sys storage

Reports mounted filesystems with usage as a percentage, which is the quickest way to find a filesystem that is filling up.

rvc sys storage

The output is a table with a row per filesystem.

 Device       Mount Point           Filesystem  Total Size  Used
-----------------------------------------------------------------
 /dev/dm-1    /                     ext4               92G   11%
 /dev/sda1    /boot                 ext2              236M   37%
 /dev/drbd50  /kvm/kvm_cluster_50   ext4               25G    1%

A device named /dev/drbdN is a replicated volume rather than local storage. Its mount point tells you which resource it belongs to. For what the DRBD device itself is doing, see DRBD replication status reference.

Configuring shell access: the account settings and SSH key needed to run these commands.
Verifying cluster health: establishing whether a cluster is healthy, end to end.
DRBD replication status reference: reading replication state and the DRBD status commands.
Reading logs on a node: local log paths, filtering, and the rsyslog spool.
Managing nodes: the node management tasks available in SC//Reliant Platform Manager.