Search <book_title>...

Storage Foundation for Oracle® RAC 7.3.1 Administrator's Guide - Linux

Last Published: 2018-01-16

Product(s): InfoScale & Storage Foundation (7.3.1)

Section I. SF Oracle RAC concepts and administration
1. Overview of Storage Foundation for Oracle RAC
2. Administering SF Oracle RAC and its components
Section II. Performance and troubleshooting
Section III. Reference
1. Appendix A. List of SF Oracle RAC health checks
2. Appendix B. Error messages

About LLT timer tunable parameters

Table: LLT timer tunable parameters lists the LLT timer tunable parameters. The timer values are set in .01 sec units. The command lltconfig - T query can be used to display current timer values.

Table: LLT timer tunable parameters

LLT parameter	Description	Default	When to change	Dependency with other LLT tunable parameters
peerinact	LLT marks a link of a peer node as "inactive," if it does not receive any packet on that link for this timer interval. Once a link is marked as "inactive," LLT will not send any data on that link.	1600	Change this value for delaying or speeding up node/link inactive notification mechanism as per client's notification processing logic. Increase the value for planned replacement of faulty network cable /switch. In some circumstances, when the private networks links are very slow or the network traffic becomes very bursty, increase this value so as to avoid false notifications of peer death. Set the value to a high value for planned replacement of faulty network cable or faulty switch.	The timer value should always be higher than the peertrouble timer value.
rpeerinact	Mark RDMA channel of a RDMA link as "inactive", if the node does not receive any packet on that link for this timer interval. Once RDMA channel is marked as "inactive", LLT does not send any data on the RDMA channel of that link, however, it may continue to send data over non-RDMA channel of that link until peerinact expires. You can view the status of the RDMA channel of a RDMA link using lltstat -nvv -r command. This parameter is supported only on selected versions of Linux.	700	Decrease the value of this tunable for speeding up the RDMA link failure recovery. If the links are unstable, and they are going up and down frequently then do not decrease this value.	This timer value should always be greater than peertrouble timer value and less than peerinact value.
peertrouble	LLT marks a high-pri link of a peer node as "troubled", if it does not receive any packet on that link for this timer interval. Once a link is marked as "troubled", LLT will not send any data on that link till the link is up.	200	In some circumstances, when the private networks links are very slow or nodes in the cluster are very busy, increase the value. Increase the value for planned replacement of faulty network cable /faulty switch.	This timer value should always be lower than peerinact timer value. Also, It should be close to its default value.
peertroublelo	LLT marks a low-pri link of a peer node as "troubled", if it does not receive any packet on that link for this timer interval. Once a link is marked as "troubled", LLT will not send any data on that link till the link is available.	400	In some circumstances, when the private networks links are very slow or nodes in the cluster are very busy, increase the value. Increase the value for planned replacement of faulty network cable /faulty switch.	This timer value should always be lower than peerinact timer value. Also, It should be close to its default value.
heartbeat	LLT sends heartbeat packets repeatedly to peer nodes after every heartbeat timer interval on each highpri link.	50	In some circumstances, when the private networks links are very slow (or congested) or nodes in the cluster are very busy, increase the value.	This timer value should be lower than peertrouble timer value. Also, it should not be close to peertrouble timer value.
heartbeatlo	LLT sends heartbeat packets repeatedly to peer nodes after every heartbeatlo timer interval on each low pri link.	100	In some circumstances, when the networks links are very slow or nodes in the cluster are very busy, increase the value.	This timer value should be lower than peertroublelo timer value. Also, it should not be close to peertroublelo timer value.
timetoreqhb	If LLT does not receive any packet from the peer node on a particular link for "timetoreqhb" time period, it attempts to request heartbeats (sends 5 special heartbeat requests (hbreqs) to the peer node on the same link) from the peer node. If the peer node does not respond to the special heartbeat requests, LLT marks the link as "expired" for that peer node. The value can be set from the range of 0 to (peerinact -200). The value 0 disables the request heartbeat mechanism.	1400	Decrease the value of this tunable for speeding up node/link inactive notification mechanism as per client's notification processing logic. Disable the request heartbeat mechanism by setting the value of this timer to 0 for planned replacement of faulty network cable /switch. In some circumstances, when the private networks links are very slow or the network traffic becomes very bursty, don't change the value of this timer tunable.	This timer is set to 'peerinact - 200' automatically every time when the peerinact timer is changed.
reqhbtime	This value specifies the time interval between two successive special heartbeat requests. See the timetoreqhb parameter for more information on special heartbeat requests.	40	Veritas recommends that you do not change this value.	Not applicable
timetosendhb	LLT sends out of timer context heartbeats to keep the node alive when LLT timer does not run at regular interval. This option specifies the amount of time to wait before sending a heartbeat in case of timer not running. If this timer tunable is set to 0, the out of timer context heartbeating mechanism is disabled.	200	Disable the out of timer context heart-beating mechanism by setting the value of this timer to 0 for planned replacement of faulty network cable /switch. In some circumstances, when the private networks links are very slow or nodes in the cluster are very busy, increase the value	This timer value should not be more than peerinact timer value. Also, it should not be close to the peerinact timer value.
sendhbcap	This value specifies the maximum time for which LLT will send contiguous out of timer context heartbeats.	18000	Veritas recommends that you do not change this value.	NA
oos	If the out-of-sequence timer has expired for a node, LLT sends an appropriate NAK to that node. LLT does not send a NAK as soon as it receives an oos packet. It waits for the oos timer value before sending the NAK.	10	Do not change this value for performance reasons. Lowering the value can result in unnecessary retransmissions/negative acknowledgement traffic. You can increase the value of oos if the round trip time is large in the cluster (for example, campus cluster).	Not applicable
retrans	LLT retransmits a packet if it does not receive its acknowledgement for this timer interval value.	10	Do not change this value. Lowering the value can result in unnecessary retransmissions. You can increase the value of retrans if the round trip time is large in the cluster (for example, campus cluster).	Not applicable
service	LLT calls its service routine (which delivers messages to LLT clients) after every service timer interval.	100	Do not change this value for performance reasons.	Not applicable
arp	LLT flushes stored address of peer nodes when this timer expires and relearns the addresses.	0	This feature is disabled by default.	Not applicable
arpreq	LLT sends an arp request when this timer expires to detect other peer nodes in the cluster.	3000	Do not change this value for performance reasons.	Not applicable