Package org.apache.cassandra.config
Class Config
java.lang.Object
org.apache.cassandra.config.Config
A class that contains configuration properties for the cassandra node it runs within.
Properties declared as volatile can be mutated via JMX.
-
Nested Class Summary
Nested ClassesModifier and TypeClassDescriptionstatic enumstatic enumstatic enumstatic enumstatic enumWhich timestamp should be used to represent a base for replica-side timeouts.static enumstatic enumstatic enumstatic enumstatic enumstatic enumstatic classstatic enumA safety mechanism for detecting incorrect paxos state, that may be down either to a bug or incorrect usage of LWTs (most likely due to unsafe mixing of SERIAL and LOCAL_SERIAL operations), and rejectingstatic enumSelect the kind of paxos state purging to use.static enumThe variants of paxos implementation and semantics supported by Cassandra.static enumstatic classstatic enum -
Field Summary
FieldsModifier and TypeFieldDescriptionTriggers automatic allocation of tokens if set, using the replication strategy of the referenced keyspaceTriggers automatic allocation of tokens if set, based on the provided replica count for a datacenterbooleanSet this to allow UDFs accessing java.lang.System.* methods, which basically allows UDFs to execute any arbitrary code on the system.booleanbooleanSet this to true to allow running insecure UDFs.booleanbooleanbooleanbooleanbooleanbooleanbooleanbooleanWhen auto_snapshot is true and this property is set, snapshots created by truncation or drop use this TTL.booleanbooleanintbooleanDeprecated.See CASSANDRA-15375Deprecated.See CASSANDRA-15375booleanintWhen a node first starts up it intially considers all other peers as DOWN and is disconnected from all of them.booleanDeprecated.See CASSANDRA-15358booleanbooleanbooleanbooleanbooleanIf these are enabled duplicate keys will get logged, and if snapshot_on_duplicate_row_detection is enabled, the table will get snapshotted for offline investigationbooleanintintintbooleanintintintintintintintintintintbooleanintintintString[]Default compaction configuration, used if a table does not specify any.intbooleanSince the denylist in many ways serves to protect the health of the cluster from partitions operators have identified as being in a bad state, we usually want more robustness than just CL.ONE on operations to/from these tables to ensure that these safeguards are in place.intWe cap the number of denylisted keys allowed per table to keep things from growing unbounded.intWe cap the total number of denylisted keys allowed in the cluster to keep things from growing unbounded.booleanbooleanbooleanbooleandoubledoublebooleanbooleanbooleanbooleanbooleanbooleandoublebooleanintintbooleanBecause of the currentBufferPoolslab sizes of 64 KiB, we store in the file cache buffers that divide 64 KiB, so we need to round the buffer sizes to powers of two.booleanbooleanIf we encounter a Gossip bug whereGossiper.getMinVersion(long, java.util.concurrent.TimeUnit)is unable to accurately report a minimum version for the cluster, optionally force the optimized index status format added in CASSANDRA-20058.booleanbooleanbooleanMaintain statistics on whether writes achieve the ideal consistency level before expiring and becoming hintsintintbooleanintThe intial capacity for creating RangeTombstoneList.booleanbooleanbooleanbooleanbooleanintintbooleanintbooleanintintbooleanbooleanThe directory to use for storing the system keyspaces data.booleanbooleanbooleanintintintintintDeprecated.See CASSANDRA-17378intintMax size of values in SSTables, in MebiBytes.intintThe configuration of timestamp boundsintlongintintbooleanbooleanintdo bcrypt hashing in a limited pool to prevent cpu load spikes; note: any value < 1 will be set to 1 on initlonglongintintintDeprecated.See CASSANDRA-19392doublebooleanbooleanbooleanintDeprecated.static final intDeprecated.intDeprecated.Deprecated.intDeprecated.static final intDeprecated.intintbooleanThis feature allows denying access to operations on certain key partitions, intended for use by operators to provide another tool to manage cluster health vs application access.intintlonglongA safety margin when purging paxos state information that has been safely replicated to a quorum.booleanEnable/disable paxos repair.intThe number of keys we may simultaneously attempt to finish incomplete paxos operations for.booleanSeeConfig.PaxosStatePurgingdefault is legacy.booleanIf true, paxos topology change repair only requires a global quorum of live nodes.booleanIf true, a quorum will be required for the global and local quorum checks.SeeConfig.PaxosVariant.booleanintdoubleSize of the CQL prepared statements cache in MiB.static final StringdoubleThe growth factor to enlarge a RangeTombstoneList.booleanbooleanbooleanintfinal RepairConfigintdoubleDeprecated.Seeintbooleanbooleanflags for enabling tracking repaired state of data during reads separate flags for range & single partition reads as single partition reads are only tracked when CL > 1 and a digest mismatch occurs.booleanbooleanintintbooleanbooleanintintbooleanbooleanDeprecated.See CASSANDRA-18252booleanintintdoublebooleanbooleanIf true, paxos topology change repair will not run on a topology change - this option should only be used in rare operation circumstances e.g.If necessary for operational purposes, permit certain keyspaces to be ignored for paxos topology repairsbooleanbooleanlongbooleanIf true, when rows with duplicate clustering keys are detected during a read or compaction a snapshot will be taken.booleanintfinal Config.SSTableConfigbooleanbooleanThe configuration of startup checks.intbooleanbooleanintintintintbooleanbooleanbooleanbooleanbooleanbooleanbooleanbooleanDeprecated.booleanbooleanbooleanTime in milliseconds after a fatal UDF run-time situation is detected and action according to user_function_timeout_policy will take place.booleanOptionally disable asynchronous UDF execution.Time in milliseconds after a warning will be emitted to the log and to the client that a UDF runs too long.Defines what to do when a UDF ran longer than user_defined_functions_fail_timeout.booleanbooleanNumber of seconds to set nowInSec into the future when performing validation previews against repaired data this (attempts) to prevent a race where validations on different machines are started on different sides of a tombstone being compacted awayintintbooleanintDeprecated.See CASSANDRA-17404booleanboolean -
Constructor Summary
Constructors -
Method Summary
Modifier and TypeMethodDescriptionstatic booleanDeprecated.migrate toDatabaseDescriptor.isClientInitialized()See CASSANDRA-12550static voidstatic voidsetClientMode(boolean clientMode) Deprecated.migrate toDatabaseDescriptor.clientInitialization(boolean)See CASSANDRA-12550static voidsetOverrideLoadConfig(Supplier<Config> loadConfig)
-
Field Details
-
PROPERTY_PREFIX
- See Also:
-
cluster_name
-
authenticator
-
authorizer
-
role_manager
-
crypto_provider
-
network_authorizer
-
cidr_authorizer
-
permissions_validity
-
permissions_cache_max_entries
public volatile int permissions_cache_max_entries -
permissions_update_interval
-
permissions_cache_active_update
public volatile boolean permissions_cache_active_update -
roles_validity
-
roles_cache_max_entries
public volatile int roles_cache_max_entries -
roles_update_interval
-
roles_cache_active_update
public volatile boolean roles_cache_active_update -
credentials_validity
-
credentials_cache_max_entries
public volatile int credentials_cache_max_entries -
credentials_update_interval
-
credentials_cache_active_update
public volatile boolean credentials_cache_active_update -
partitioner
-
auto_bootstrap
public boolean auto_bootstrap -
hinted_handoff_enabled
public volatile boolean hinted_handoff_enabled -
hinted_handoff_disabled_datacenters
-
max_hint_window
-
hints_directory
-
hint_window_persistent_enabled
public boolean hint_window_persistent_enabled -
force_new_prepared_statement_behaviour
public volatile boolean force_new_prepared_statement_behaviour -
seed_provider
-
disk_access_mode
-
disk_failure_policy
-
commit_failure_policy
-
use_deterministic_table_id
Deprecated. -
initial_token
-
num_tokens
-
allocate_tokens_for_keyspace
Triggers automatic allocation of tokens if set, using the replication strategy of the referenced keyspace -
allocate_tokens_for_local_replication_factor
Triggers automatic allocation of tokens if set, based on the provided replica count for a datacenter -
native_transport_idle_timeout
-
request_timeout
-
read_request_timeout
-
range_request_timeout
-
write_request_timeout
-
counter_write_request_timeout
-
cas_contention_timeout
-
truncate_request_timeout
-
repair_request_timeout
-
streaming_connections_per_host
-
streaming_keep_alive_period
-
internode_timeout
public boolean internode_timeout -
slow_query_log_timeout
-
stream_transfer_task_timeout
-
phi_convict_threshold
public volatile double phi_convict_threshold -
concurrent_reads
public int concurrent_reads -
concurrent_writes
public int concurrent_writes -
concurrent_counter_writes
public int concurrent_counter_writes -
concurrent_materialized_view_writes
public int concurrent_materialized_view_writes -
available_processors
public int available_processors -
memtable_flush_writers
public int memtable_flush_writers -
memtable_heap_space
-
memtable_offheap_space
-
memtable_cleanup_threshold
-
memtable
-
repair_session_max_tree_depth
Deprecated.See -
repair_session_space
-
concurrent_merkle_tree_requests
public volatile int concurrent_merkle_tree_requests -
use_offheap_merkle_trees
public volatile boolean use_offheap_merkle_trees -
storage_port
public int storage_port -
ssl_storage_port
public int ssl_storage_port -
listen_address
-
listen_interface
-
listen_interface_prefer_ipv6
public boolean listen_interface_prefer_ipv6 -
broadcast_address
-
listen_on_broadcast_address
public boolean listen_on_broadcast_address -
internode_authenticator
-
traverse_auth_from_root
public boolean traverse_auth_from_root -
rpc_address
-
rpc_interface
-
rpc_interface_prefer_ipv6
public boolean rpc_interface_prefer_ipv6 -
broadcast_rpc_address
-
rpc_keepalive
public boolean rpc_keepalive -
internode_max_message_size
-
internode_socket_send_buffer_size
-
internode_socket_receive_buffer_size
-
internode_application_send_queue_capacity
-
internode_application_send_queue_reserve_endpoint_capacity
-
internode_application_send_queue_reserve_global_capacity
-
internode_application_receive_queue_capacity
-
internode_application_receive_queue_reserve_endpoint_capacity
-
internode_application_receive_queue_reserve_global_capacity
-
internode_tcp_connect_timeout
-
internode_tcp_user_timeout
-
internode_streaming_tcp_user_timeout
-
start_native_transport
public boolean start_native_transport -
native_transport_port
public int native_transport_port -
native_transport_port_ssl
Deprecated.See CASSANDRA-19392 -
native_transport_max_threads
public int native_transport_max_threads -
native_transport_max_frame_size
-
native_transport_max_auth_threads
public int native_transport_max_auth_threadsdo bcrypt hashing in a limited pool to prevent cpu load spikes; note: any value < 1 will be set to 1 on init -
native_transport_max_message_size
-
native_transport_max_concurrent_connections
public volatile long native_transport_max_concurrent_connections -
native_transport_max_concurrent_connections_per_ip
public volatile long native_transport_max_concurrent_connections_per_ip -
native_transport_flush_in_batches_legacy
public boolean native_transport_flush_in_batches_legacy -
native_transport_allow_older_protocols
public volatile boolean native_transport_allow_older_protocols -
native_transport_max_request_data_in_flight_per_ip
-
native_transport_max_request_data_in_flight
-
native_transport_rate_limiting_enabled
public volatile boolean native_transport_rate_limiting_enabled -
native_transport_max_requests_per_second
public volatile int native_transport_max_requests_per_second -
native_transport_receive_queue_capacity
-
max_value_size
Max size of values in SSTables, in MebiBytes. Default is the same as the native protocol frame limit: 256MiB. See AbstractType for how it is used. -
snapshot_before_compaction
public boolean snapshot_before_compaction -
auto_snapshot
public boolean auto_snapshot -
auto_snapshot_ttl
When auto_snapshot is true and this property is set, snapshots created by truncation or drop use this TTL. -
snapshot_links_per_second
public volatile long snapshot_links_per_second -
column_index_size
-
column_index_cache_size
-
batch_size_warn_threshold
-
batch_size_fail_threshold
-
unlogged_batch_across_partitions_warn_threshold
-
concurrent_compactors
-
compaction_throughput
-
min_free_space_per_drive
-
compressed_read_ahead_buffer_size
-
max_space_usable_for_compactions_in_percentage
-
concurrent_materialized_view_builders
public volatile int concurrent_materialized_view_builders -
reject_repair_compaction_threshold
public volatile int reject_repair_compaction_threshold -
concurrent_index_builders
public volatile int concurrent_index_builders -
repair_disk_headroom_reject_ratio
public volatile double repair_disk_headroom_reject_ratio -
max_streaming_retries
Deprecated.See CASSANDRA-17378 -
stream_throughput_outbound
-
inter_dc_stream_throughput_outbound
-
entire_sstable_stream_throughput_outbound
-
entire_sstable_inter_dc_stream_throughput_outbound
public volatile DataRateSpec.LongBytesPerSecondBound entire_sstable_inter_dc_stream_throughput_outbound -
data_file_directories
-
sstable
-
local_system_data_file_directory
The directory to use for storing the system keyspaces data. If unspecified the data will be stored in the first of the data_file_directories. -
saved_caches_directory
-
commitlog_directory
-
commitlog_total_space
-
commitlog_sync
-
commitlog_sync_group_window
-
commitlog_sync_period
-
commitlog_segment_size
-
commitlog_compression
-
flush_compression
-
commitlog_max_compression_buffers_in_pool
public int commitlog_max_compression_buffers_in_pool -
commitlog_disk_access_mode
-
periodic_commitlog_sync_lag_block
-
transparent_data_encryption_options
-
max_mutation_size
-
cdc_enabled
public boolean cdc_enabled -
cdc_block_writes
public volatile boolean cdc_block_writes -
cdc_on_repair_enabled
public volatile boolean cdc_on_repair_enabled -
cdc_raw_directory
-
cdc_total_space
-
cdc_free_space_check_interval
-
endpoint_snitch
-
dynamic_snitch
public boolean dynamic_snitch -
dynamic_snitch_update_interval
-
dynamic_snitch_reset_interval
-
dynamic_snitch_badness_threshold
public double dynamic_snitch_badness_threshold -
failure_detector
-
server_encryption_options
-
client_encryption_options
-
internode_compression
-
hinted_handoff_throttle
-
batchlog_replay_throttle
-
batchlog_endpoint_strategy
-
max_hints_delivery_threads
public int max_hints_delivery_threads -
hints_flush_period
-
max_hints_file_size
-
max_hints_size_per_host
-
hints_compression
-
auto_hints_cleanup_enabled
public volatile boolean auto_hints_cleanup_enabled -
transfer_hints_on_decommission
public volatile boolean transfer_hints_on_decommission -
use_creation_time_for_hint_ttl
public volatile boolean use_creation_time_for_hint_ttl -
incremental_backups
public volatile boolean incremental_backups -
trickle_fsync
public boolean trickle_fsync -
trickle_fsync_interval
-
sstable_preemptive_open_interval
-
key_cache_migrate_during_compaction
public volatile boolean key_cache_migrate_during_compaction -
key_cache_invalidate_after_sstable_deletion
public volatile boolean key_cache_invalidate_after_sstable_deletion -
key_cache_keys_to_save
public volatile int key_cache_keys_to_save -
key_cache_size
-
key_cache_save_period
-
row_cache_class_name
-
row_cache_size
-
row_cache_save_period
-
row_cache_keys_to_save
public volatile int row_cache_keys_to_save -
counter_cache_size
-
counter_cache_save_period
-
counter_cache_keys_to_save
public volatile int counter_cache_keys_to_save -
paxos_cache_size
-
cache_load_timeout
-
networking_cache_size
-
file_cache_size
-
file_cache_enabled
public boolean file_cache_enabled -
file_cache_round_up
Because of the currentBufferPoolslab sizes of 64 KiB, we store in the file cache buffers that divide 64 KiB, so we need to round the buffer sizes to powers of two. This boolean controls weather they are rounded up or down. Set it to true to round up to the next power of two, set it to false to round down to the previous power of two. Note that buffer sizes are already rounded to 4 KiB and capped between 4 KiB minimum and 64 kb maximum by theConfig.DiskOptimizationStrategy. By default, this boolean is set to round down whendisk_optimization_strategyisssd, and to round up when it isspinning. -
buffer_pool_use_heap_if_exhausted
Deprecated.See CASSANDRA-15358 -
disk_optimization_strategy
-
disk_optimization_estimate_percentile
public double disk_optimization_estimate_percentile -
disk_optimization_page_cross_chance
public double disk_optimization_page_cross_chance -
inter_dc_tcp_nodelay
public boolean inter_dc_tcp_nodelay -
memtable_allocation_type
-
read_thresholds_enabled
public volatile boolean read_thresholds_enabled -
coordinator_read_size_warn_threshold
-
coordinator_read_size_fail_threshold
-
local_read_size_warn_threshold
-
local_read_size_fail_threshold
-
row_index_read_size_warn_threshold
-
row_index_read_size_fail_threshold
-
tombstone_warn_threshold
public volatile int tombstone_warn_threshold -
tombstone_failure_threshold
public volatile int tombstone_failure_threshold -
replica_filtering_protection
-
index_summary_capacity
-
index_summary_resize_interval
-
gc_log_threshold
-
gc_warn_threshold
-
trace_type_query_ttl
-
trace_type_repair_ttl
-
ideal_consistency_level
Maintain statistics on whether writes achieve the ideal consistency level before expiring and becoming hints -
windows_timer_interval
Deprecated.See CASSANDRA-17404 -
otc_coalescing_strategy
Deprecated. -
otc_coalescing_window_us_default
Deprecated.- See Also:
-
otc_coalescing_window_us
Deprecated. -
otc_coalescing_enough_coalesced_messages
Deprecated. -
otc_backlog_expiration_interval_ms_default
Deprecated.- See Also:
-
otc_backlog_expiration_interval_ms
Deprecated. -
prepared_statements_cache_size
Size of the CQL prepared statements cache in MiB. Defaults to 1/256th of the heap size or 10MiB, whichever is greater. -
user_defined_functions_enabled
public boolean user_defined_functions_enabled -
scripted_user_defined_functions_enabled
Deprecated.See CASSANDRA-18252 -
materialized_views_enabled
public boolean materialized_views_enabled -
materialized_views_on_repair_enabled
public volatile boolean materialized_views_on_repair_enabled -
transient_replication_enabled
public boolean transient_replication_enabled -
sasi_indexes_enabled
public boolean sasi_indexes_enabled -
drop_compact_storage_enabled
public volatile boolean drop_compact_storage_enabled -
use_statements_enabled
public volatile boolean use_statements_enabled -
user_defined_functions_threads_enabled
public boolean user_defined_functions_threads_enabledOptionally disable asynchronous UDF execution. Disabling asynchronous UDF execution also implicitly disables the security-manager! By default, async UDF execution is enabled to be able to detect UDFs that run too long / forever and be able to fail fast - i.e. stop the Cassandra daemon, which is currently the only appropriate approach to "tell" a user that there's something really wrong with the UDF. When you disable async UDF execution, users MUST pay attention to read-timeouts since these may indicate UDFs that run too long or forever - and this can destabilize the cluster. This requires allow_insecure_udfs to be true -
allow_insecure_udfs
public boolean allow_insecure_udfsSet this to true to allow running insecure UDFs. -
allow_extra_insecure_udfs
public boolean allow_extra_insecure_udfsSet this to allow UDFs accessing java.lang.System.* methods, which basically allows UDFs to execute any arbitrary code on the system. -
dynamic_data_masking_enabled
public boolean dynamic_data_masking_enabled -
user_defined_functions_warn_timeout
Time in milliseconds after a warning will be emitted to the log and to the client that a UDF runs too long. (Only valid, if user_defined_functions_threads_enabled==true) -
user_defined_functions_fail_timeout
Time in milliseconds after a fatal UDF run-time situation is detected and action according to user_function_timeout_policy will take place. (Only valid, if user_defined_functions_threads_enabled==true) -
user_function_timeout_policy
Defines what to do when a UDF ran longer than user_defined_functions_fail_timeout. Possible options are: - 'die' - i.e. it is able to emit a warning to the client before the Cassandra Daemon will shut down. - 'die_immediate' - shut down C* daemon immediately (effectively prevent the chance that the client will receive a warning). - 'ignore' - just log - the most dangerous option. (Only valid, if user_defined_functions_threads_enabled==true) -
back_pressure_enabled
Deprecated.See CASSANDRA-15375 -
back_pressure_strategy
Deprecated.See CASSANDRA-15375 -
concurrent_validations
public volatile int concurrent_validations -
repair_command_pool_full_strategy
-
repair_command_pool_size
public int repair_command_pool_size -
block_for_peers_timeout_in_secs
public int block_for_peers_timeout_in_secsWhen a node first starts up it intially considers all other peers as DOWN and is disconnected from all of them. To be useful as a coordinator (and not introduce latency penalties on restart) this node must have successfully opened all three internode TCP connections (gossip, small, and large messages) before advertising to clients. Due to this, by default, Casssandra will prime these internode TCP connections and wait for all but a single node to be DOWN/disconnected in the local datacenter before offering itself as a coordinator, subject to a timeout. See CASSANDRA-13993 and CASSANDRA-14297 for more details. We provide two tunables to control this behavior as some users may want to block until all datacenters are available (global QUORUM/EACH_QUORUM), some users may not want to block at all (clients that already work around the problem), and some users may want to prime the connections but not delay startup. block_for_peers_timeout_in_secs: controls how long this node will wait to connect to peers. To completely disable any startup connectivity checks set this to -1. To trigger the internode connections but immediately continue startup, set this to to 0. The default is 10 seconds. block_for_peers_in_remote_dcs: controls if this node will consider remote datacenters to wait for. The default is to _not_ wait on remote datacenters. -
block_for_peers_in_remote_dcs
public boolean block_for_peers_in_remote_dcs -
automatic_sstable_upgrade
public volatile boolean automatic_sstable_upgrade -
max_concurrent_automatic_sstable_upgrades
public volatile int max_concurrent_automatic_sstable_upgrades -
stream_entire_sstables
public boolean stream_entire_sstables -
skip_stream_disk_space_check
public volatile boolean skip_stream_disk_space_check -
audit_logging_options
-
full_query_logging_options
-
corrupted_tombstone_strategy
-
diagnostic_events_enabled
public volatile boolean diagnostic_events_enabled -
default_keyspace_rf
public volatile int default_keyspace_rf -
repaired_data_tracking_for_range_reads_enabled
public volatile boolean repaired_data_tracking_for_range_reads_enabledflags for enabling tracking repaired state of data during reads separate flags for range & single partition reads as single partition reads are only tracked when CL > 1 and a digest mismatch occurs. Currently, range queries don't use digests so if enabled for range reads, all such reads will include repaired data tracking. As this adds some overhead, operators may wish to disable it whilst still enabling it for partition reads -
repaired_data_tracking_for_partition_reads_enabled
public volatile boolean repaired_data_tracking_for_partition_reads_enabled -
report_unconfirmed_repaired_data_mismatches
public volatile boolean report_unconfirmed_repaired_data_mismatches -
snapshot_on_repaired_data_mismatch
public volatile boolean snapshot_on_repaired_data_mismatch -
validation_preview_purge_head_start
Number of seconds to set nowInSec into the future when performing validation previews against repaired data this (attempts) to prevent a race where validations on different machines are started on different sides of a tombstone being compacted away -
auth_cache_warming_enabled
public boolean auth_cache_warming_enabled -
auth_read_consistency_level
-
auth_write_consistency_level
-
partition_denylist_enabled
public volatile boolean partition_denylist_enabledThis feature allows denying access to operations on certain key partitions, intended for use by operators to provide another tool to manage cluster health vs application access. See CASSANDRA-12106 and CEP-13 for more details. -
denylist_writes_enabled
public volatile boolean denylist_writes_enabled -
denylist_reads_enabled
public volatile boolean denylist_reads_enabled -
denylist_range_reads_enabled
public volatile boolean denylist_range_reads_enabled -
denylist_refresh
-
denylist_initial_load_retry
-
denylist_max_keys_per_table
public volatile int denylist_max_keys_per_tableWe cap the number of denylisted keys allowed per table to keep things from growing unbounded. Operators will receive warnings and only denylist_max_keys_per_table in natural query ordering will be processed on overflow. -
denylist_max_keys_total
public volatile int denylist_max_keys_totalWe cap the total number of denylisted keys allowed in the cluster to keep things from growing unbounded. Operators will receive warnings on initial cache load that there are too many keys and be directed to trim down the entries to within the configured limits. -
denylist_consistency_level
Since the denylist in many ways serves to protect the health of the cluster from partitions operators have identified as being in a bad state, we usually want more robustness than just CL.ONE on operations to/from these tables to ensure that these safeguards are in place. That said, we allow users to configure this if they're so inclined. -
log_out_of_token_range_requests
public volatile boolean log_out_of_token_range_requests -
reject_out_of_token_range_requests
public volatile boolean reject_out_of_token_range_requests -
initial_range_tombstone_list_allocation_size
public volatile int initial_range_tombstone_list_allocation_sizeThe intial capacity for creating RangeTombstoneList. -
range_tombstone_list_growth_factor
public volatile double range_tombstone_list_growth_factorThe growth factor to enlarge a RangeTombstoneList. -
sai_options
-
snapshot_on_duplicate_row_detection
public volatile boolean snapshot_on_duplicate_row_detectionIf true, when rows with duplicate clustering keys are detected during a read or compaction a snapshot will be taken. In the read case, each a snapshot request will be issued to each replica involved in the query, for compaction the snapshot will be created locally. These are limited at the replica level so that only a single snapshot per-day can be taken via this method. This requires check_for_duplicate_rows_during_reads and/or check_for_duplicate_rows_during_compaction below to be enabled -
check_for_duplicate_rows_during_reads
public volatile boolean check_for_duplicate_rows_during_readsIf these are enabled duplicate keys will get logged, and if snapshot_on_duplicate_row_detection is enabled, the table will get snapshotted for offline investigation -
check_for_duplicate_rows_during_compaction
public volatile boolean check_for_duplicate_rows_during_compaction -
autocompaction_on_startup_enabled
public boolean autocompaction_on_startup_enabled -
auto_optimise_inc_repair_streams
public volatile boolean auto_optimise_inc_repair_streams -
auto_optimise_full_repair_streams
public volatile boolean auto_optimise_full_repair_streams -
auto_optimise_preview_repair_streams
public volatile boolean auto_optimise_preview_repair_streams -
uuid_sstable_identifiers_enabled
public boolean uuid_sstable_identifiers_enabled -
consecutive_message_errors_threshold
public volatile int consecutive_message_errors_threshold -
client_error_reporting_exclusions
-
internode_error_reporting_exclusions
-
invalid_legacy_protocol_magic_no_spam_enabled
public volatile boolean invalid_legacy_protocol_magic_no_spam_enabled -
keyspaces_warn_threshold
public volatile int keyspaces_warn_threshold -
keyspaces_fail_threshold
public volatile int keyspaces_fail_threshold -
tables_warn_threshold
public volatile int tables_warn_threshold -
tables_fail_threshold
public volatile int tables_fail_threshold -
columns_per_table_warn_threshold
public volatile int columns_per_table_warn_threshold -
columns_per_table_fail_threshold
public volatile int columns_per_table_fail_threshold -
secondary_indexes_per_table_warn_threshold
public volatile int secondary_indexes_per_table_warn_threshold -
secondary_indexes_per_table_fail_threshold
public volatile int secondary_indexes_per_table_fail_threshold -
materialized_views_per_table_warn_threshold
public volatile int materialized_views_per_table_warn_threshold -
materialized_views_per_table_fail_threshold
public volatile int materialized_views_per_table_fail_threshold -
page_size_warn_threshold
public volatile int page_size_warn_threshold -
page_size_fail_threshold
public volatile int page_size_fail_threshold -
partition_keys_in_select_warn_threshold
public volatile int partition_keys_in_select_warn_threshold -
partition_keys_in_select_fail_threshold
public volatile int partition_keys_in_select_fail_threshold -
in_select_cartesian_product_warn_threshold
public volatile int in_select_cartesian_product_warn_threshold -
in_select_cartesian_product_fail_threshold
public volatile int in_select_cartesian_product_fail_threshold -
table_properties_warned
-
table_properties_ignored
-
table_properties_disallowed
-
read_consistency_levels_warned
-
read_consistency_levels_disallowed
-
write_consistency_levels_warned
-
write_consistency_levels_disallowed
-
user_timestamps_enabled
public volatile boolean user_timestamps_enabled -
alter_table_enabled
public volatile boolean alter_table_enabled -
group_by_enabled
public volatile boolean group_by_enabled -
drop_truncate_table_enabled
public volatile boolean drop_truncate_table_enabled -
drop_keyspace_enabled
public volatile boolean drop_keyspace_enabled -
secondary_indexes_enabled
public volatile boolean secondary_indexes_enabled -
force_optimized_index_status_format
public volatile boolean force_optimized_index_status_formatIf we encounter a Gossip bug whereGossiper.getMinVersion(long, java.util.concurrent.TimeUnit)is unable to accurately report a minimum version for the cluster, optionally force the optimized index status format added in CASSANDRA-20058. -
default_secondary_index
-
default_secondary_index_enabled
public volatile boolean default_secondary_index_enabled -
uncompressed_tables_enabled
public volatile boolean uncompressed_tables_enabled -
compact_tables_enabled
public volatile boolean compact_tables_enabled -
read_before_write_list_operations_enabled
public volatile boolean read_before_write_list_operations_enabled -
allow_filtering_enabled
public volatile boolean allow_filtering_enabled -
simplestrategy_enabled
public volatile boolean simplestrategy_enabled -
partition_size_warn_threshold
-
partition_size_fail_threshold
-
partition_tombstones_warn_threshold
public volatile long partition_tombstones_warn_threshold -
partition_tombstones_fail_threshold
public volatile long partition_tombstones_fail_threshold -
column_value_size_warn_threshold
-
column_value_size_fail_threshold
-
collection_size_warn_threshold
-
collection_size_fail_threshold
-
items_per_collection_warn_threshold
public volatile int items_per_collection_warn_threshold -
items_per_collection_fail_threshold
public volatile int items_per_collection_fail_threshold -
fields_per_udt_warn_threshold
public volatile int fields_per_udt_warn_threshold -
fields_per_udt_fail_threshold
public volatile int fields_per_udt_fail_threshold -
vector_type_enabled
public volatile boolean vector_type_enabled -
vector_dimensions_warn_threshold
public volatile int vector_dimensions_warn_threshold -
vector_dimensions_fail_threshold
public volatile int vector_dimensions_fail_threshold -
data_disk_usage_percentage_warn_threshold
public volatile int data_disk_usage_percentage_warn_threshold -
data_disk_usage_percentage_fail_threshold
public volatile int data_disk_usage_percentage_fail_threshold -
data_disk_usage_max_disk_size
-
minimum_replication_factor_warn_threshold
public volatile int minimum_replication_factor_warn_threshold -
minimum_replication_factor_fail_threshold
public volatile int minimum_replication_factor_fail_threshold -
maximum_replication_factor_warn_threshold
public volatile int maximum_replication_factor_warn_threshold -
maximum_replication_factor_fail_threshold
public volatile int maximum_replication_factor_fail_threshold -
zero_ttl_on_twcs_warned
public volatile boolean zero_ttl_on_twcs_warned -
zero_ttl_on_twcs_enabled
public volatile boolean zero_ttl_on_twcs_enabled -
non_partition_restricted_index_query_enabled
public volatile boolean non_partition_restricted_index_query_enabled -
intersect_filtering_query_warned
public volatile boolean intersect_filtering_query_warned -
intersect_filtering_query_enabled
public volatile boolean intersect_filtering_query_enabled -
sai_sstable_indexes_per_query_warn_threshold
public volatile int sai_sstable_indexes_per_query_warn_threshold -
sai_sstable_indexes_per_query_fail_threshold
public volatile int sai_sstable_indexes_per_query_fail_threshold -
sai_string_term_size_warn_threshold
-
sai_string_term_size_fail_threshold
-
sai_frozen_term_size_warn_threshold
-
sai_frozen_term_size_fail_threshold
-
sai_vector_term_size_warn_threshold
-
sai_vector_term_size_fail_threshold
-
streaming_state_expires
-
streaming_state_size
-
streaming_stats_enabled
public volatile boolean streaming_stats_enabled -
streaming_slow_events_log_timeout
-
startup_checks
The configuration of startup checks. -
repair_state_expires
-
repair_state_size
public volatile int repair_state_size -
maximum_timestamp_warn_threshold
The configuration of timestamp bounds -
maximum_timestamp_fail_threshold
-
minimum_timestamp_warn_threshold
-
minimum_timestamp_fail_threshold
-
auto_repair
-
paxos_variant
SeeConfig.PaxosVariant. Defaults to v1, recommend upgrading to v2 at earliest opportunity. -
skip_paxos_repair_on_topology_change
public volatile boolean skip_paxos_repair_on_topology_changeIf true, paxos topology change repair will not run on a topology change - this option should only be used in rare operation circumstances e.g. where for some reason the repair is impossible to perform (e.g. too few replicas) and an unsafe topology change must be made -
paxos_purge_grace_period
A safety margin when purging paxos state information that has been safely replicated to a quorum. Data for transactions initiated within this grace period will not be expunged. -
paxos_on_linearizability_violations
SeeConfig.PaxosOnLinearizabilityViolation. Default is to ignore, as applications may readily mix SERIAL and LOCAL_SERIAL and this is the most likely source of linearizability violations. this facility should be activated only for debugging Cassandra or by power users who are investigating their own application behaviour. -
paxos_state_purging
SeeConfig.PaxosStatePurgingdefault is legacy. -
paxos_repair_enabled
public volatile boolean paxos_repair_enabledEnable/disable paxos repair. This is a global flag that not only determines default behaviour but overrides explicit paxos repair requests, paxos repair on topology changes and paxos auto repairs. -
paxos_topology_repair_no_dc_checks
public boolean paxos_topology_repair_no_dc_checksIf true, paxos topology change repair only requires a global quorum of live nodes. If false, it requires a global quorum as well as a local quorum for each dc (EACH_QUORUM), with the exception explained in paxos_topology_repair_strict_each_quorum -
paxos_topology_repair_strict_each_quorum
public boolean paxos_topology_repair_strict_each_quorumIf true, a quorum will be required for the global and local quorum checks. If false, we will accept a quorum OR n - 1 live nodes. This is to allow for topologies like 2:2:2, where paxos queries always use SERIAL, and a single node down in a dc should not preclude a paxos repair -
skip_paxos_repair_on_topology_change_keyspaces
If necessary for operational purposes, permit certain keyspaces to be ignored for paxos topology repairs -
paxos_contention_wait_randomizer
-
paxos_contention_min_wait
-
paxos_contention_max_wait
-
paxos_contention_min_delta
-
paxos_repair_parallelism
public volatile int paxos_repair_parallelismThe number of keys we may simultaneously attempt to finish incomplete paxos operations for. -
sstable_read_rate_persistence_enabled
public volatile boolean sstable_read_rate_persistence_enabled -
client_request_size_metrics_enabled
public volatile boolean client_request_size_metrics_enabled -
max_top_size_partition_count
public volatile int max_top_size_partition_count -
max_top_tombstone_partition_count
public volatile int max_top_tombstone_partition_count -
min_tracked_partition_size
-
min_tracked_partition_tombstone_count
public volatile long min_tracked_partition_tombstone_count -
top_partitions_enabled
public volatile boolean top_partitions_enabled -
repair
-
default_compaction
Default compaction configuration, used if a table does not specify any. -
dump_heap_on_uncaught_exception
public volatile boolean dump_heap_on_uncaught_exception -
heap_dump_path
-
severity_during_decommission
public double severity_during_decommission -
storage_compatibility_mode
-
cql_start_time
-
native_transport_throw_on_overload
public boolean native_transport_throw_on_overload -
native_transport_queue_max_item_age_threshold
public double native_transport_queue_max_item_age_threshold -
native_transport_min_backoff_on_queue_overload
-
native_transport_max_backoff_on_queue_overload
-
native_transport_timeout
-
enforce_native_deadline_for_hints
public boolean enforce_native_deadline_for_hints -
paxos_repair_race_wait
public boolean paxos_repair_race_wait
-
-
Constructor Details
-
Config
public Config()
-
-
Method Details
-
splitCommaDelimited
-
isClientMode
Deprecated.migrate toDatabaseDescriptor.isClientInitialized()See CASSANDRA-12550 -
setClientMode
Deprecated.migrate toDatabaseDescriptor.clientInitialization(boolean)See CASSANDRA-12550Client mode means that the process is a pure client, that uses C* code base but does not read or write local C* database files. -
getOverrideLoadConfig
-
setOverrideLoadConfig
-
log
-