LVS Load Balancing command description and round robin explanation of IPVSADM

LVS Load Balancing command description and round robin explanation of IPVSADM

"1" common parameters of IPVSADM:
-A: Add A new virtual server record to the virtual server table of the kernel.
-E: edit a virtual server record in the kernel virtual server table.
-D: delete a virtual server record in the kernel virtual server table.
-C: Clear all records in the kernel virtual server table
-R: Restores virtual server rules.
-S: saves the virtual server rules and outputs the readable format of the-R option.
-A: Add a new Real Server record to a record in the kernel virtual server table.
-E: edit a Real Server record in a virtual server record.
-D: delete a Real Server record in a virtual server record.
-L |-l: displays the kernel virtual server table.
-Z: virtual service table counters are cleared (the current number of connections is cleared, etc)
-Set:-tcp tcpfin udp sets the connection timeout value
-- Start-daemon: start the synchronization daemon process.
-- Stop-daemon: stops the synchronization daemon process.
-H: displays help information.
-T: indicates that the virtual server provides the TCP Service.
-U: indicates that the virtual server provides the UDP Service.
-F: indicates the service type marked by iptables.
-S: scheduling algorithm used. Common options include rr | wrr | lc | wlc | lblc | lblcr | dh | sh | sed | nq
-P: Persistent Service
-R: Real Server
-G: Specify the LVS working mode as direct routing mode.
-I: Specify the LVS working mode as tunnel mode.
-M: Specify the LVS working mode as NAT
-W: actual server weight
-C: displays the number of current LVS connections.
-Timeout: display the timeout value of tcp tcpfin udp
-- Daemon: displays the synchronization daemon status.
-- Stats: displays statistics.
-- Rate: Display rate information
-- Sort: sorting and outputting virtual servers and real servers
-N: Number of output IP addresses and ports

"2" Round Robin explanation:
RR: The pure round robin method, which is junk. Distribute each request to the Real Server in order.
WRR: Weighted Round Robin (WRR. Distribute each request cyclically in the Real Server in order, but assign more jobs to the servers with higher capacities.
LC: distribute based on the minimum number of connections.
WLC: weighted. The weight configured on the machine is high.
LBLC: cache server cluster. Based on the local minimum connection. Transfer the request to a server with a low load.
LBLCR: cache server cluster with replication scheduling. When A page is cached on server A, it is frequently accessed. When other cache servers have low load, it monitors whether to access the same page. If the same page is accessed, the request is distributed to another server.
DH: Target hash scheduling, which is bound to two ip addresses in realserver. Ld identifies the ISP provider of the new ISP and forwards it to the corresponding IP address.
SH: Source Address hash. Client address-based source differentiation. (Rarely used)
SED: the shortest expected latency.
NQ: Minimum queue scheduling, no queue required. If the number of connections of a realserver is equal to 0, it is directly allocated. sed is not required.

This article permanently updates link: https://www.bkjia.com/Linux/2018-03/151515.htm