Advanced Troubleshooting; Enabling Power On After Power Has Been Lost; Vlan Networking Problems; Software Update Process - HPE StoreOnce 6500 User Manual

Advanced troubleshooting

More detailed troubleshooting information can be found at www.hpe.com/support/hpesc.

Enabling power on after power has been lost

In the event of a cold reboot when power has been lost to the whole system, the order in which power is
re-applied is important. The recommended and preferred order for switching on the circuit breakers to the
PDU is in the following sequence (often from the top down, but the location of circuit breakers varies,
depending upon the type of PDU). This enables power to the pre-installed hardware in the following
sequence:
1. Network switches will power on immediately on connection to the PDU. Check for fault lights on the
switches (front and rear of the rack).
2. Power on the disk enclosures using the power button (on the upper left as seen from the rear).
3. After 10 minutes, power on the nodes (servers) using the power button on the front of each server.

VLAN networking problems

The StoreOnce System supports VLAN tagging.
The following guidelines apply:
• The VLAN tagging interfaces can be configured on both single and bonded interfaces.
• The user can configure a maximum of 128 IPv4 VLANs and 128 IPv6 VLANs per appliance; they may
be consecutive or non-consecutive.
• The user can configure all VLANs on one Port Set, or divide the VLANs over multiple Port Sets.
• Each VLAN tag interface is configured using the VLAN ID index, which is the same as the switch
VLAN ID number and must be an integer from 2 to 4094.
• Each VLAN tag interface can be configured for use with one IPv4 and one IPv6 subnet. For example,
if a VLAN tag ID was configured for use with ports eth 0 and eth1 on IPv4, it can also be configured for
use on IPv6 with that Port Set but it cannot be used with any other Port Set.

Software update process

Software updates can only be performed through the StoreOnce CLI using the administrator user
account.
Software updates are delivered to customers in the form of a single rpm package, which may contain
multiple rpms. Everything is updated in the same way, using the rpm package: the software, third-party
components and the firmware for hardware components.
All nodes are updated in parallel. A software update is NOT an online process. Services will shut down to
perform the updates, so the whole appliance will be offline for a period of time (up to 60 minutes).
IMPORTANT:
A software upgrade process cannot be performed on a cluster if one or more nodes are down. In
this case, the node needs to be brought up and working prior to the software update process.
Similarly, if a service set is faulty, the upgrade may not work in some circumstances
Advanced troubleshooting 241