Veritas Storage Foundation 5.1 SP1 for Oracle RAC Administrator"s Guide (5900-1512, April 2011)

Table 3-17
Troubleshooting LLT warning messages
RecommendationPossible causesWarning
Check for applications that
may be throttling the CPU
and memory resources on the
node.
CPU and memory consumption
on the node is high.
Warning: OS timer
is not called for
num seconds
Reduce memory
consumption on the node and
free up allocated memory.
The available memory on the
node is insufficient.
Warning: Kernel
failed to allocate
memory num time(s)
■ Allocate more bandwidth
for communication
between the local node
and the peer node.
■ Check for applications
that may be throttling the
CPU and memory
resources on the node.
■ The network bandwidth
between the local node and
the peer node (node number)
is insufficient.
■ The CPU and memory
consumption on the peer node
(node number) is very high.
Flow-control occurred
num time(s)and
back-enabled num
time(s) on port port
number for node node
number
Check the connectivity
between the local node and
the peer node. Replace the
link, if needed.
The private interconnect between
the local node and the peer node
is unstable.
Warning:
Connectivity with
node node id on
link link id is
flaky num time(s).
The distribution
is: (0-4 s) <num>
(4-8 s) <num> (8-12
s) <num> (12-16 s)
<num> (>=16 s)
Check the private
interconnects between the
local node and the peer node.
The private interconnects
between the local node and peer
node may not have sufficient
bandwidth.
One or more link
connectivity with
peer node(s) node
name is in trouble.
■ If the private
interconnect is faulty,
replace the link.
■ Configure a minimum of
two private
interconnects.
■ One of the configured private
interconnects is
non-operational.
■ Only one private interconnect
has been configured under
LLT.
Link connectivity
with node id is on
only one link.
Symantec recommends
configuring a
minimum of 2 links.
Troubleshooting SF Oracle RAC
Troubleshooting LLT health check warning messages
198