Coroutines,advised to set the value to twice the threads
server.max_procs
2
-
Threads
server.db_path
-
-
Storage directory
server.slow_time
"30ms"
-
Time threshold of slow query
server.slow_key_window_time
"2000ms"
-
The cumulative slow query time for single key in statistical window. If cumulative time exceeds threshold, the request of related key will be forbidden.
server.slow_shield
true
-
Whether to enable slow query forbidden
server.slow_ttl
1s
-
Statistical window for slow query
server.slow_maxexec
100
-
The maximum number of concurrent slow queries allowed
server.slow_topn
100
-
Forbid top-n slow query only
server.open_distributed_tx
false
false
Whether to enable transactions, Bitalosproxy needs to be configured collaboratively置
The compact start time of bitalosdb, an hour in 24 hours
bitalos.compact_end_time
6
6
The compact end time of bitalosdb, an hour in 24 hours
bitalos.compact_interval
300
300
Bitalosdb compact interval
bitalos.bithash_gc_threshold
0.5
0.5
Bitalosdb's Bithash invalid data proportion threshold, compact when reaching the threshold
bitalos.compression_type
0
0
Bitalosdb compression type of bitalosdb, 0=no compression, 1=snappy compression
bitalos.enable_expired_deletion
true
true
Whether to enable deletion of expired data
bitalos.expired_deletion_interval
60
60
Interval of delete expired data
bitalos.expired_deletion_qps_threshold
20000
20000
Expired data deletion qps threshold. When qps is greater than the threshold, expired data will not be deleted.
bitalos.expired_deletion_disable_start_time
0
0
Expired data deletion time (start), expired data deletion is not performed from expired_deletion_disable_start_time to expired_deletion_disable_end_time
bitalos.expired_deletion_disable_end_time
0
0
Expired data deletion time (end), expired data deletion is not performed from expired_deletion_disable_start_time to expired_deletion_disable_end_time
bitalos.io_write_qps_threshold
20000
20000
IO control threshold, used for bitalosdb's IOWriteLoadThresholdFunc callback to determine whether qps exceeds the threshold. If it exceeds, return false, otherwise return true
bitalos.max_field_size
10240
10240
Maximum length of of field for composite data type (set/zset/list/hash), default value is 10240, maximum value is 61440
bitalos.max_value_size
6291456
6291456
Maximum length of value, default value is 6291456, maximum value is 8388608
bitalos.enable_raftlog_restore
false
false
Whether to enable raftlog recovery. If enabled, it is recommended to set bitalos.enable_wal false
bitalos.enable_page_block_compression
false
false
Whether to enable Bitpage block compression，suggest to set true for storing huge small values
bitalos.enable_clock_cache
false
true
Whether to enable clock cache. Clock cache period is 1 second. Set it to false if time accuracy is within milliseconds.
raft_queue.workers
60
60
Gorountine number of consumimg raft message for slave node
raft_queue.length
10000
10000
Maximum length of single consume goroutine for slave node
raft_cluster.cluster_id
-
-
raft cluster id
raft_cluster.election_rtt
35
-
Trigger election cycle
raft_cluster.heartbeat_rtt
1
-
Heartbeat cycle
raft_cluster.check_quorm
true
true
Whether master node periodically checks the majority node status. If the majority node status is not obtained, master will automatically become slave node.
raft_cluster.snapshot_entries
6000000
-
Raftlog snapshot threshold, raftlog will be cleaned when the threshold is exceeded
Raft operation timeout, such as adding/removing node, propose logs, etc.
raft_cluster.retry_times
1
-
Number of retries to add/remove raft nodes
raft_cluster.async_propose
true
-
Whether to enable raftlog asynchronous proposal
raft_cluster.is_observer
-
-
Is observer node or not
raft_cluster.is_witness
-
-
Is witness node or not
raft_nodehost.host_name
-
-
Machine hostname
raft_nodehost.node_id
-
-
Raft node id
raft_nodehost.raft_address
-
-
raft service address
raft_nodehost.init_raft_addrlist
-
-
raftinitial node address list
raft_nodehost.init_raft_nodelist
-
-
raft initial node id list
raft_nodehost.join
false
-
raft_nodehost.snapshot_timeout
"10s"
-
snapshot timeout
raft_nodehost.rtt
200
-
Heartbeat rtt, unit: milliseconds
raft_nodehost.deployment_id
-
-
Deployment id, default is 0
raft_nodehost.max_snapshot_send_bytes_persecod
"60mb"
-
Number of bytes per second to send snapshot segment
raft_nodehost.max_snapshot_recv_bytes_persecod
"60mb"
-
Number of bytes per second to receive snapshot segment
raft_state.interval
"3s"
-
Periodically detect whether the latest heartbeat of the storage node is within the election cycle. If it exceeds the election cycle, the node status becomes false.
dynamic_deadline.client_ratio_threshold
[0,20,50,80,90]
-
Ratio of service connection to max connection (max_clients))
dynamic_deadline.deadline_threshold
["1800s","600s","180s","60s","10s"]
-
According to current service connection number, set corresponding connect timeout for all connections. For example, client_ratio_threshold is 30%~59%, connection timeout is 100s.
