IBM FHIR Server 4.5.0 - Performance

Workloads

Ingestion
Client Access

Ingestion Workload

The fhir-bucket utility scans an IBM Cloud Object Store (COS) bucket for matching json or ndjson files. Matching files are recorded in the fhirbucket database and this information is subsequently used to drive the loading of those files into the LinuxForHealth FHIR Server.

The test described here uses bundles based on Synthea - Synthetic Mass data. The original bundles have been preprocessed and broken into smaller bundles of no more than 100 resources. Resources within original bundles contain references to other resources within the same bundle. Traditionally, these references would be resolved by the LinuxForHealth FHIR Server on ingestion. However, because the bundles are being broken into smaller units of work, the references must be resolved in advance. This requires that any create requests using server-asserted ids must be rewritten as client asserted id requests, using PUT.

The fhir-bucket uses a large thread-pool to fetch the bundles from COS and POST each to the LinuxForHealth FHIR Server. The result bundle is parsed and each of the returned logical resource ids are stored in the fhirbucket database, along with some timing metrics.

Client Access

The client access workload is generated using fhir-bucket running within the same K8s cluster. Random patients are selected from the fhirbucket.logical_resources table which is populated by first running the fhir-bucket ingestion mode. The client workload is comprised of the following steps:

Select a random patient id and submit to the thread-pool;
Create a Bundle resource containing a read for the patient resource and a search for ExplanationOfBenefit resources which refer to that patient limit 20;
POST the bundle request;
Process the response, constructing a new Bundle resource containing GET requests for each Claim and Provider referenced by the ExplanationOfBenefit resources contained in the first response.
Collect statistics on the number of calls, number of resources returned and the time taken for each request.

The request bundles are configured to use a type of TRANSACTION. This provides the best performance as the LinuxForHealth FHIR Server processes each request within the bundle using a single transaction, reducing the commit load on the database.

System Configuration

Configuration Item	Value
IBM FHIR Server	Release 4.5.0
Client thread-pool size	300
FHIR Server Deployment Size	6 PODS
PostgreSQL Service	IBM Cloud Database
PostgreSQL Version	12.4
PostgreSQL CPU	9 cores dedicated
PostgreSQL RAM	24GB
PostgreSQL Storage	1TB up to 10K IOPS

Top Resource Counts

Top resource counts in the database using the query:

  SELECT rt.resource_type, count(*)
    FROM fhirdata.logical_resources lr,
         fhirdata.resource_types rt
   WHERE rt.resource_type_id = lr.resource_type_id
GROUP BY rt.resource_type
ORDER BY 2 DESC;

resource_type	count
Observation	28643640
Claim	9809811
Procedure	9533768
Encounter	8788209
ExplanationOfBenefit	7795369
DiagnosticReport	2942204
MedicationRequest	2761835
Condition	1755335
Immunization	1656535
CarePlan	1320978
Goal	347835
CareTeam	341226
Practitioner	319783
Organization	319729
AllergyIntolerance	216350
Patient	174388

The full list is included at the end of this report.

Ingestion Performance

To compute the ingestion rate we can query the fhirbucket.logical_resources table, looking for the first and last creation times for the latest run of the loader, loader_instance_id = 212 in the example below:

select count(*), max(created_tstamp) - min(created_tstamp) as elapsed 
  from fhirbucket.logical_resources lr, fhirbucket.resource_bundle_loads bl 
 where bl.resource_bundle_load_id = lr.resource_bundle_load_id
   and bl.loader_instance_id = 212;

count	elapsed
525525	00:09:19.253087

Ingestion rate:

525525 resources 
---------------- = 940 resources/second
   559 seconds

The inserts into the Observation parameter tables for token_values and quantity_values come from composite values, and there is room for improvement for how these inserts are handled:

The GC activity is very reasonable given the complexity of FHIR resource parsing and validation:

The throughput is stable throughout the loading period:

The top database operations reflect what is seen in the application breakdown, with observation_token_values and observation_quantity_values being a good target for future improvement:

Time series view of the same data:

Client Read Performance

The average server response time reported by the client is consistently under 180ms as shown below (statistics are sampled and printed by fhir-bucket every 10 seconds:

STATS: FHIR= 1643.8 calls/s, rate=16825.1 resources/s, response time=0.179 s
STATS: FHIR= 1655.6 calls/s, rate=16860.8 resources/s, response time=0.178 s
STATS: FHIR= 1693.2 calls/s, rate=17323.5 resources/s, response time=0.176 s
STATS: FHIR= 1699.1 calls/s, rate=17431.5 resources/s, response time=0.174 s
STATS: FHIR= 1647.1 calls/s, rate=16849.2 resources/s, response time=0.179 s
STATS: FHIR= 1676.6 calls/s, rate=17180.6 resources/s, response time=0.176 s
STATS: FHIR= 1691.3 calls/s, rate=17242.2 resources/s, response time=0.175 s
STATS: FHIR= 1663.7 calls/s, rate=17043.8 resources/s, response time=0.178 s
STATS: FHIR= 1639.9 calls/s, rate=16723.9 resources/s, response time=0.175 s

Throughput is constrained by exhaustion of PostgreSQL CPU.

Client Workload Performance Metrics

Database processing rate is slightly more than 3000 transactions per second. This reached 15K transactions/second when using BATCH as the bundle type instead of TRANSACTION. In that case, although the TPS is higher, the overall throughput is lower because each request required multiple transactions.

The peak aggregate CPU usage for the POD (including the fhirbucket instance driving the test workload) is around 36 cores.

No significant hotspots in the application breakdown:

GC times are remarkably good, especially given the complexity of processing FHIR resources.

The top DB operations align with the above Application Breakdown.

The database CPU is hitting its limit, suggesting that scaling up to more cores would increase throughput.

Postgres Database Configuration

Name	Category	Setting	Unit	Description
DateStyle	Client Connection Defaults / Locale and Formatting	ISO, MDY		Sets the display format for date and time values.
IntervalStyle	Client Connection Defaults / Locale and Formatting	postgres		Sets the display format for interval values.
TimeZone	Client Connection Defaults / Locale and Formatting	UTC		Sets the time zone for displaying and interpreting time stamps.
allow_system_table_mods	Developer Options	off		Allows modifications of the structure of system tables.
application_name	Reporting and Logging / What to Log	pgAdmin 4 - DB:fhirdb		Sets the application name to be reported in statistics and logs.
archive_cleanup_command	Write-Ahead Log / Archive Recovery			Sets the shell command that will be executed at every restart point.
archive_command	Write-Ahead Log / Archiving	/usr/bin/pgbackrest —config /conf/postgresql/pgbackrest.conf —stanza=formation archive-push %p		Sets the shell command that will be called to archive a WAL file.
archive_mode	Write-Ahead Log / Archiving	on		Allows archiving of WAL files using archive_command.
archive_timeout	Write-Ahead Log / Archiving	1800	s	Forces a switch to the next WAL file if a new file has not been started within N seconds.
array_nulls	Version and Platform Compatibility / Previous PostgreSQL Versions	on		Enable input of NULL elements in arrays.
authentication_timeout	Connections and Authentication / Authentication	30	s	Sets the maximum allowed time to complete client authentication.
autovacuum	Autovacuum	on		Starts the autovacuum subprocess.
autovacuum_analyze_scale_factor	Autovacuum	0.1		Number of tuple inserts, updates, or deletes prior to analyze as a fraction of reltuples.
autovacuum_analyze_threshold	Autovacuum	50		Minimum number of tuple inserts, updates, or deletes prior to analyze.
autovacuum_freeze_max_age	Autovacuum	200000000		Age at which to autovacuum a table to prevent transaction ID wraparound.
autovacuum_max_workers	Autovacuum	3		Sets the maximum number of simultaneously running autovacuum worker processes.
autovacuum_multixact_freeze_max_age	Autovacuum	400000000		Multixact age at which to autovacuum a table to prevent multixact wraparound.
autovacuum_naptime	Autovacuum	60	s	Time to sleep between autovacuum runs.
autovacuum_vacuum_cost_delay	Autovacuum	20	ms	Vacuum cost delay in milliseconds, for autovacuum.
autovacuum_vacuum_cost_limit	Autovacuum	-1		Vacuum cost amount available before napping, for autovacuum.
autovacuum_vacuum_scale_factor	Autovacuum	0.2		Number of tuple updates or deletes prior to vacuum as a fraction of reltuples.
autovacuum_vacuum_threshold	Autovacuum	50		Minimum number of tuple updates or deletes prior to vacuum.
autovacuum_work_mem	Resource Usage / Memory	-1	kB	Sets the maximum memory to be used by each autovacuum worker process.
backend_flush_after	Resource Usage / Asynchronous Behavior	0	8kB	Number of pages after which previously performed writes are flushed to disk.
backslash_quote	Version and Platform Compatibility / Previous PostgreSQL Versions	safe_encoding		Sets whether ”\’” is allowed in string literals.
bgwriter_delay	Resource Usage / Background Writer	200	ms	Background writer sleep time between rounds.
bgwriter_flush_after	Resource Usage / Background Writer	64	8kB	Number of pages after which previously performed writes are flushed to disk.
bgwriter_lru_maxpages	Resource Usage / Background Writer	100		Background writer maximum number of LRU pages to flush per round.
bgwriter_lru_multiplier	Resource Usage / Background Writer	2		Multiple of the average buffer usage to free per round.
block_size	Preset Options	8192		Shows the size of a disk block.
bonjour	Connections and Authentication / Connection Settings	off		Enables advertising the server via Bonjour.
bonjour_name	Connections and Authentication / Connection Settings			Sets the Bonjour service name.
bytea_output	Client Connection Defaults / Statement Behavior	hex		Sets the output format for bytea.
check_function_bodies	Client Connection Defaults / Statement Behavior	on		Check function bodies during CREATE FUNCTION.
checkpoint_completion_target	Write-Ahead Log / Checkpoints	0.7		Time spent flushing dirty buffers during checkpoint, as fraction of checkpoint interval.
checkpoint_flush_after	Write-Ahead Log / Checkpoints	32	8kB	Number of pages after which previously performed writes are flushed to disk.
checkpoint_timeout	Write-Ahead Log / Checkpoints	300	s	Sets the maximum time between automatic WAL checkpoints.
checkpoint_warning	Write-Ahead Log / Checkpoints	30	s	Enables warnings if checkpoint segments are filled more frequently than this.
client_encoding	Client Connection Defaults / Locale and Formatting	UNICODE		Sets the client’s character set encoding.
client_min_messages	Client Connection Defaults / Statement Behavior	notice		Sets the message levels that are sent to the client.
cluster_name	Process Title	postgresql		Sets the name of the cluster, which is included in the process title.
commit_delay	Write-Ahead Log / Settings	0		Sets the delay in microseconds between transaction commit and flushing WAL to disk.
commit_siblings	Write-Ahead Log / Settings	5		Sets the minimum concurrent open transactions before performing commit_delay.
constraint_exclusion	Query Tuning / Other Planner Options	partition		Enables the planner to use constraints to optimize queries.
cpu_index_tuple_cost	Query Tuning / Planner Cost Constants	0.005		Sets the planner’s estimate of the cost of processing each index entry during an index scan.
cpu_operator_cost	Query Tuning / Planner Cost Constants	0.0025		Sets the planner’s estimate of the cost of processing each operator or function call.
cpu_tuple_cost	Query Tuning / Planner Cost Constants	0.01		Sets the planner’s estimate of the cost of processing each tuple (row).
cursor_tuple_fraction	Query Tuning / Other Planner Options	0.1		Sets the planner’s estimate of the fraction of a cursor’s rows that will be retrieved.
data_checksums	Preset Options	on		Shows whether data checksums are turned on for this cluster.
data_directory_mode	Preset Options	0700		Mode of the data directory.
data_sync_retry	Error Handling	off		Whether to continue running after a failure to sync data files.
db_user_namespace	Connections and Authentication / Authentication	off		Enables per-database user names.
deadlock_timeout	Lock Management	10000	ms	Sets the time to wait on a lock before checking for deadlock.
debug_assertions	Preset Options	off		Shows whether the running server has assertion checks enabled.
debug_pretty_print	Reporting and Logging / What to Log	on		Indents parse and plan tree displays.
debug_print_parse	Reporting and Logging / What to Log	off		Logs each query’s parse tree.
debug_print_plan	Reporting and Logging / What to Log	off		Logs each query’s execution plan.
debug_print_rewritten	Reporting and Logging / What to Log	off		Logs each query’s rewritten parse tree.
default_statistics_target	Query Tuning / Other Planner Options	100		Sets the default statistics target.
default_table_access_method	Client Connection Defaults / Statement Behavior	heap		Sets the default table access method for new tables.
default_tablespace	Client Connection Defaults / Statement Behavior			Sets the default tablespace to create tables and indexes in.
default_text_search_config	Client Connection Defaults / Locale and Formatting	pg_catalog.english		Sets default text search configuration.
default_transaction_deferrable	Client Connection Defaults / Statement Behavior	off		Sets the default deferrable status of new transactions.
default_transaction_isolation	Client Connection Defaults / Statement Behavior	read committed		Sets the transaction isolation level of each new transaction.
default_transaction_read_only	Client Connection Defaults / Statement Behavior	off		Sets the default read-only status of new transactions.
dynamic_shared_memory_type	Resource Usage / Memory	posix		Selects the dynamic shared memory implementation used.
effective_cache_size	Query Tuning / Planner Cost Constants	2359296	8kB	Sets the planner’s assumption about the total size of the data caches.
effective_io_concurrency	Resource Usage / Asynchronous Behavior	12		Number of simultaneous requests that can be handled efficiently by the disk subsystem.
enable_bitmapscan	Query Tuning / Planner Method Configuration	on		Enables the planner’s use of bitmap-scan plans.
enable_gathermerge	Query Tuning / Planner Method Configuration	on		Enables the planner’s use of gather merge plans.
enable_hashagg	Query Tuning / Planner Method Configuration	on		Enables the planner’s use of hashed aggregation plans.
enable_hashjoin	Query Tuning / Planner Method Configuration	on		Enables the planner’s use of hash join plans.
enable_indexonlyscan	Query Tuning / Planner Method Configuration	on		Enables the planner’s use of index-only-scan plans.
enable_indexscan	Query Tuning / Planner Method Configuration	on		Enables the planner’s use of index-scan plans.
enable_material	Query Tuning / Planner Method Configuration	on		Enables the planner’s use of materialization.
enable_mergejoin	Query Tuning / Planner Method Configuration	on		Enables the planner’s use of merge join plans.
enable_nestloop	Query Tuning / Planner Method Configuration	on		Enables the planner’s use of nested-loop join plans.
enable_parallel_append	Query Tuning / Planner Method Configuration	on		Enables the planner’s use of parallel append plans.
enable_parallel_hash	Query Tuning / Planner Method Configuration	on		Enables the planner’s use of parallel hash plans.
enable_partition_pruning	Query Tuning / Planner Method Configuration	on		Enables plan-time and run-time partition pruning.
enable_partitionwise_aggregate	Query Tuning / Planner Method Configuration	off		Enables partitionwise aggregation and grouping.
enable_partitionwise_join	Query Tuning / Planner Method Configuration	off		Enables partitionwise join.
enable_seqscan	Query Tuning / Planner Method Configuration	on		Enables the planner’s use of sequential-scan plans.
enable_sort	Query Tuning / Planner Method Configuration	on		Enables the planner’s use of explicit sort steps.
enable_tidscan	Query Tuning / Planner Method Configuration	on		Enables the planner’s use of TID scan plans.
escape_string_warning	Version and Platform Compatibility / Previous PostgreSQL Versions	on		Warn about backslash escapes in ordinary string literals.
event_source	Reporting and Logging / Where to Log	PostgreSQL		Sets the application name used to identify PostgreSQL messages in the event log.
exit_on_error	Error Handling	off		Terminate session on any error.
extra_float_digits	Client Connection Defaults / Locale and Formatting	1		Sets the number of digits displayed for floating-point values.
force_parallel_mode	Query Tuning / Other Planner Options	off		Forces use of parallel query facilities.
from_collapse_limit	Query Tuning / Other Planner Options	8		Sets the FROM-list size beyond which subqueries are not collapsed.
fsync	Write-Ahead Log / Settings	on		Forces synchronization of updates to disk.
full_page_writes	Write-Ahead Log / Settings	on		Writes full pages to WAL when first modified after a checkpoint.
geqo	Query Tuning / Genetic Query Optimizer	on		Enables genetic query optimization.
geqo_effort	Query Tuning / Genetic Query Optimizer	5		GEQO: effort is used to set the default for other GEQO parameters.
geqo_generations	Query Tuning / Genetic Query Optimizer	0		GEQO: number of iterations of the algorithm.
geqo_pool_size	Query Tuning / Genetic Query Optimizer	0		GEQO: number of individuals in the population.
geqo_seed	Query Tuning / Genetic Query Optimizer	0		GEQO: seed for random path selection.
geqo_selection_bias	Query Tuning / Genetic Query Optimizer	2		GEQO: selective pressure within the population.
geqo_threshold	Query Tuning / Genetic Query Optimizer	12		Sets the threshold of FROM items beyond which GEQO is used.
gin_fuzzy_search_limit	Client Connection Defaults / Other Defaults	0		Sets the maximum allowed result for exact search by GIN.
gin_pending_list_limit	Client Connection Defaults / Statement Behavior	4096	kB	Sets the maximum size of the pending list for GIN index.
hot_standby	Replication / Standby Servers	on		Allows connections and queries during recovery.
hot_standby_feedback	Replication / Standby Servers	off		Allows feedback from a hot standby to the primary that will avoid query conflicts.
huge_pages	Resource Usage / Memory	off		Use of huge pages on Linux or Windows.
idle_in_transaction_session_timeout	Client Connection Defaults / Statement Behavior	0	ms	Sets the maximum allowed duration of any idling transaction.
ignore_checksum_failure	Developer Options	off		Continues processing after a checksum failure.
ignore_system_indexes	Developer Options	off		Disables reading from system indexes.
integer_datetimes	Preset Options	on		Datetimes are integer based.
jit	Query Tuning / Other Planner Options	on		Allow JIT compilation.
jit_above_cost	Query Tuning / Planner Cost Constants	100000		Perform JIT compilation if query is more expensive.
jit_debugging_support	Developer Options	off		Register JIT compiled function with debugger.
jit_dump_bitcode	Developer Options	off		Write out LLVM bitcode to facilitate JIT debugging.
jit_expressions	Developer Options	on		Allow JIT compilation of expressions.
jit_inline_above_cost	Query Tuning / Planner Cost Constants	500000		Perform JIT inlining if query is more expensive.
jit_optimize_above_cost	Query Tuning / Planner Cost Constants	500000		Optimize JITed functions if query is more expensive.
jit_profiling_support	Developer Options	off		Register JIT compiled function with perf profiler.
jit_tuple_deforming	Developer Options	on		Allow JIT compilation of tuple deforming.
join_collapse_limit	Query Tuning / Other Planner Options	8		Sets the FROM-list size beyond which JOIN constructs are not flattened.
krb_caseins_users	Connections and Authentication / Authentication	off		Sets whether Kerberos and GSSAPI user names should be treated as case-insensitive.
lc_collate	Client Connection Defaults / Locale and Formatting	C.UTF-8		Shows the collation order locale.
lc_ctype	Client Connection Defaults / Locale and Formatting	C.UTF-8		Shows the character classification and case conversion locale.
lc_messages	Client Connection Defaults / Locale and Formatting	en_US.UTF-8		Sets the language in which messages are displayed.
lc_monetary	Client Connection Defaults / Locale and Formatting	en_US.UTF-8		Sets the locale for formatting monetary amounts.
lc_numeric	Client Connection Defaults / Locale and Formatting	en_US.UTF-8		Sets the locale for formatting numbers.
lc_time	Client Connection Defaults / Locale and Formatting	en_US.UTF-8		Sets the locale for formatting date and time values.
listen_addresses	Connections and Authentication / Connection Settings	0.0.0.0		Sets the host name or IP address(es) to listen to.
lo_compat_privileges	Version and Platform Compatibility / Previous PostgreSQL Versions	off		Enables backward compatibility mode for privilege checks on large objects.
local_preload_libraries	Client Connection Defaults / Shared Library Preloading	pgextwlist		Lists unprivileged shared libraries to preload into each backend.
lock_timeout	Client Connection Defaults / Statement Behavior	0	ms	Sets the maximum allowed duration of any wait for a lock.
log_autovacuum_min_duration	Reporting and Logging / What to Log	1000	ms	Sets the minimum execution time above which autovacuum actions will be logged.
log_checkpoints	Reporting and Logging / What to Log	on		Logs each checkpoint.
log_connections	Reporting and Logging / What to Log	off		Logs each successful connection.
log_destination	Reporting and Logging / Where to Log	stderr		Sets the destination for server log output.
log_disconnections	Reporting and Logging / What to Log	off		Logs end of a session, including duration.
log_duration	Reporting and Logging / What to Log	off		Logs the duration of each completed SQL statement.
log_error_verbosity	Reporting and Logging / What to Log	default		Sets the verbosity of logged messages.
log_executor_stats	Statistics / Monitoring	off		Writes executor performance statistics to the server log.
log_file_mode	Reporting and Logging / Where to Log	0600		Sets the file permissions for log files.
log_hostname	Reporting and Logging / What to Log	off		Logs the host name in the connection logs.
log_line_prefix	Reporting and Logging / What to Log	%t [%a][%e] [%p]: [%l-1] user=%u,db=%d,client=%h		Controls information prefixed to each log line.
log_lock_waits	Reporting and Logging / What to Log	on		Logs long lock waits.
log_min_duration_statement	Reporting and Logging / When to Log	100	ms	Sets the minimum execution time above which statements will be logged.
log_min_error_statement	Reporting and Logging / When to Log	error		Causes all statements generating error at or above this level to be logged.
log_min_messages	Reporting and Logging / When to Log	notice		Sets the message levels that are logged.
log_parser_stats	Statistics / Monitoring	off		Writes parser performance statistics to the server log.
log_planner_stats	Statistics / Monitoring	off		Writes planner performance statistics to the server log.
log_replication_commands	Reporting and Logging / What to Log	off		Logs each replication command.
log_rotation_age	Reporting and Logging / Where to Log	1440	min	Automatic log file rotation will occur after N minutes.
log_rotation_size	Reporting and Logging / Where to Log	10240	kB	Automatic log file rotation will occur after N kilobytes.
log_statement	Reporting and Logging / What to Log	none		Sets the type of statements logged.
log_statement_stats	Statistics / Monitoring	off		Writes cumulative performance statistics to the server log.
log_temp_files	Reporting and Logging / What to Log	0	kB	Log the use of temporary files larger than this number of kilobytes.
log_timezone	Reporting and Logging / What to Log	UTC		Sets the time zone to use in log messages.
log_transaction_sample_rate	Reporting and Logging / When to Log	0		Set the fraction of transactions to log for new transactions.
log_truncate_on_rotation	Reporting and Logging / Where to Log	off		Truncate existing log files of same name during log rotation.
logging_collector	Reporting and Logging / Where to Log	off		Start a subprocess to capture stderr output and/or csvlogs into log files.
maintenance_work_mem	Resource Usage / Memory	1572864	kB	Sets the maximum memory to be used for maintenance operations.
max_connections	Connections and Authentication / Connection Settings	400		Sets the maximum number of concurrent connections.
max_files_per_process	Resource Usage / Kernel Resources	1000		Sets the maximum number of simultaneously open files for each server process.
max_function_args	Preset Options	100		Shows the maximum number of function arguments.
max_identifier_length	Preset Options	63		Shows the maximum identifier length.
max_index_keys	Preset Options	32		Shows the maximum number of index keys.
max_locks_per_transaction	Lock Management	64		Sets the maximum number of locks per transaction.
max_logical_replication_workers	Replication / Subscribers	4		Maximum number of logical replication worker processes.
max_parallel_maintenance_workers	Resource Usage / Asynchronous Behavior	2		Sets the maximum number of parallel processes per maintenance operation.
max_parallel_workers	Resource Usage / Asynchronous Behavior	8		Sets the maximum number of parallel workers that can be active at one time.
max_parallel_workers_per_gather	Resource Usage / Asynchronous Behavior	2		Sets the maximum number of parallel processes per executor node.
max_pred_locks_per_page	Lock Management	2		Sets the maximum number of predicate-locked tuples per page.
max_pred_locks_per_relation	Lock Management	-2		Sets the maximum number of predicate-locked pages and tuples per relation.
max_pred_locks_per_transaction	Lock Management	64		Sets the maximum number of predicate locks per transaction.
max_prepared_transactions	Resource Usage / Memory	150		Sets the maximum number of simultaneously prepared transactions.
max_replication_slots	Replication / Sending Servers	10		Sets the maximum number of simultaneously defined replication slots.
max_stack_depth	Resource Usage / Memory	2048	kB	Sets the maximum stack depth, in kilobytes.
max_standby_archive_delay	Replication / Standby Servers	30000	ms	Sets the maximum delay before canceling queries when a hot standby server is processing archived WAL data.
max_standby_streaming_delay	Replication / Standby Servers	30000	ms	Sets the maximum delay before canceling queries when a hot standby server is processing streamed WAL data.
max_sync_workers_per_subscription	Replication / Subscribers	2		Maximum number of table synchronization workers per subscription.
max_wal_senders	Replication / Sending Servers	12		Sets the maximum number of simultaneously running WAL sender processes.
max_wal_size	Write-Ahead Log / Checkpoints	1024	MB	Sets the WAL size that triggers a checkpoint.
max_worker_processes	Resource Usage / Asynchronous Behavior	8		Maximum number of concurrent worker processes.
min_parallel_index_scan_size	Query Tuning / Planner Cost Constants	64	8kB	Sets the minimum amount of index data for a parallel scan.
min_parallel_table_scan_size	Query Tuning / Planner Cost Constants	1024	8kB	Sets the minimum amount of table data for a parallel scan.
min_wal_size	Write-Ahead Log / Checkpoints	80	MB	Sets the minimum size to shrink the WAL to.
old_snapshot_threshold	Resource Usage / Asynchronous Behavior	-1	min	Time before a snapshot is too old to read pages changed after the snapshot was taken.
operator_precedence_warning	Version and Platform Compatibility / Previous PostgreSQL Versions	off		Emit a warning for constructs that changed meaning since PostgreSQL 9.4.
parallel_leader_participation	Resource Usage / Asynchronous Behavior	on		Controls whether Gather and Gather Merge also run subplans.
parallel_setup_cost	Query Tuning / Planner Cost Constants	1000		Sets the planner’s estimate of the cost of starting up worker processes for parallel query.
parallel_tuple_cost	Query Tuning / Planner Cost Constants	0.1		Sets the planner’s estimate of the cost of passing each tuple (row) from worker to master backend.
password_encryption	Connections and Authentication / Authentication	md5		Chooses the algorithm for encrypting passwords.
pg_stat_statements.max	Customized Options	5000		Sets the maximum number of statements tracked by pg_stat_statements.
pg_stat_statements.save	Customized Options	on		Save pg_stat_statements statistics across server shutdowns.
pg_stat_statements.track	Customized Options	top		Selects which statements are tracked by pg_stat_statements.
pg_stat_statements.track_utility	Customized Options	on		Selects whether utility commands are tracked by pg_stat_statements.
pgaudit.log	Customized Options	none		Specifies which classes of statements will be logged by session audit logging. Multiple classes can be provided using a comma-separated list and classes can be subtracted by prefacing the class with a - sign.
pgaudit.log_catalog	Customized Options	off		Specifies that session logging should be enabled in the case where all relations in a statement are in pg_catalog. Disabling this setting will reduce noise in the log from tools like psql and PgAdmin that query the catalog heavily.
pgaudit.log_client	Customized Options	off		Specifies whether audit messages should be visible to the client. This setting should generally be left disabled but may be useful for debugging or other purposes.
pgaudit.log_level	Customized Options	log		Specifies the log level that will be used for log entries. This setting is used for regression testing and may also be useful to end users for testing or other purposes. It is not intended to be used in a production environment as it may leak which statements are being logged to the user.
pgaudit.log_parameter	Customized Options	off		Specifies that audit logging should include the parameters that were passed with the statement. When parameters are present they will be be included in CSV format after the statement text.
pgaudit.log_relation	Customized Options	off		Specifies whether session audit logging should create a separate log entry for each relation referenced in a SELECT or DML statement. This is a useful shortcut for exhaustive logging without using object audit logging.
pgaudit.log_statement_once	Customized Options	off		Specifies whether logging will include the statement text and parameters with the first log entry for a statement/substatement combination or with every entry. Disabling this setting will result in less verbose logging but may make it more difficult to determine the statement that generated a log entry, though the statement/substatement pair along with the process id should suffice to identify the statement text logged with a previous entry.
pgaudit.role	Customized Options			Specifies the master role to use for object audit logging. Muliple audit roles can be defined by granting them to the master role. This allows multiple groups to be in charge of different aspects of audit logging.
plan_cache_mode	Query Tuning / Other Planner Options	auto		Controls the planner’s selection of custom or generic plan.
port	Connections and Authentication / Connection Settings	5432		Sets the TCP port the server listens on.
post_auth_delay	Developer Options	0	s	Waits N seconds on connection startup after authentication.
pre_auth_delay	Developer Options	0	s	Waits N seconds on connection startup before authentication.
primary_slot_name	Replication / Standby Servers			Sets the name of the replication slot to use on the sending server.
promote_trigger_file	Replication / Standby Servers			Specifies a file name whose presence ends recovery in the standby.
quote_all_identifiers	Version and Platform Compatibility / Previous PostgreSQL Versions	off		When generating SQL fragments, quote all identifiers.
random_page_cost	Query Tuning / Planner Cost Constants	2.5		Sets the planner’s estimate of the cost of a nonsequentially fetched disk page.
recovery_end_command	Write-Ahead Log / Archive Recovery			Sets the shell command that will be executed once at the end of recovery.
recovery_min_apply_delay	Replication / Standby Servers	0	ms	Sets the minimum delay for applying changes during recovery.
recovery_target	Write-Ahead Log / Recovery Target			Set to “immediate” to end recovery as soon as a consistent state is reached.
recovery_target_action	Write-Ahead Log / Recovery Target	pause		Sets the action to perform upon reaching the recovery target.
recovery_target_inclusive	Write-Ahead Log / Recovery Target	on		Sets whether to include or exclude transaction with recovery target.
recovery_target_lsn	Write-Ahead Log / Recovery Target			Sets the LSN of the write-ahead log location up to which recovery will proceed.
recovery_target_name	Write-Ahead Log / Recovery Target			Sets the named restore point up to which recovery will proceed.
recovery_target_time	Write-Ahead Log / Recovery Target			Sets the time stamp up to which recovery will proceed.
recovery_target_timeline	Write-Ahead Log / Recovery Target	latest		Specifies the timeline to recover into.
recovery_target_xid	Write-Ahead Log / Recovery Target			Sets the transaction ID up to which recovery will proceed.
restart_after_crash	Error Handling	on		Reinitialize server after backend crash.
restore_command	Write-Ahead Log / Archive Recovery	/usr/bin/pgbackrest —config /conf/postgresql/pgbackrest.conf —stanza=formation archive-get %f “%p”		Sets the shell command that will retrieve an archived WAL file.
row_security	Client Connection Defaults / Statement Behavior	on		Enable row security.
search_path	Client Connection Defaults / Statement Behavior	“$user”, public		Sets the schema search order for names that are not schema-qualified.
segment_size	Preset Options	131072	8kB	Shows the number of pages per disk file.
seq_page_cost	Query Tuning / Planner Cost Constants	1		Sets the planner’s estimate of the cost of a sequentially fetched disk page.
server_encoding	Client Connection Defaults / Locale and Formatting	UTF8		Sets the server (database) character set encoding.
server_version	Preset Options	12.4		Shows the server version.
server_version_num	Preset Options	120004		Shows the server version as an integer.
session_replication_role	Client Connection Defaults / Statement Behavior	origin		Sets the session’s behavior for triggers and rewrite rules.
shared_buffers	Resource Usage / Memory	786432	8kB	Sets the number of shared memory buffers used by the server.
shared_memory_type	Resource Usage / Memory	mmap		Selects the shared memory implementation used for the main shared memory region.
ssl	Connections and Authentication / SSL	on		Enables SSL connections.
ssl_ca_file	Connections and Authentication / SSL			Location of the SSL certificate authority file.
ssl_cert_file	Connections and Authentication / SSL	/conf/cert/tls.crt		Location of the SSL server certificate file.
ssl_crl_file	Connections and Authentication / SSL			Location of the SSL certificate revocation list file.
ssl_key_file	Connections and Authentication / SSL	/conf/cert/tls.key		Location of the SSL server private key file.
ssl_library	Preset Options	OpenSSL		Name of the SSL library.
ssl_passphrase_command	Connections and Authentication / SSL			Command to obtain passphrases for SSL.
ssl_passphrase_command_supports_reload	Connections and Authentication / SSL	off		Also use ssl_passphrase_command during server reload.
ssl_prefer_server_ciphers	Connections and Authentication / SSL	on		Give priority to server ciphersuite order.
standard_conforming_strings	Version and Platform Compatibility / Previous PostgreSQL Versions	on		Causes ’…’ strings to treat backslashes literally.
statement_timeout	Client Connection Defaults / Statement Behavior	43200000	ms	Sets the maximum allowed duration of any statement.
superuser_reserved_connections	Connections and Authentication / Connection Settings	15		Sets the number of connection slots reserved for superusers.
synchronize_seqscans	Version and Platform Compatibility / Previous PostgreSQL Versions	on		Enable synchronized sequential scans.
synchronous_commit	Write-Ahead Log / Settings	local		Sets the current transaction’s synchronization level.
synchronous_standby_names	Replication / Master Server	“c-a343dd6d-23fd-4229-aa12-1171a4fc8c13-m-0”,“c-a343dd6d-23fd-4229-aa12-1171a4fc8c13-m-1”		Number of synchronous standbys and list of names of potential synchronous ones.
syslog_facility	Reporting and Logging / Where to Log	local0		Sets the syslog “facility” to be used when syslog enabled.
syslog_ident	Reporting and Logging / Where to Log	postgres		Sets the program name used to identify PostgreSQL messages in syslog.
syslog_sequence_numbers	Reporting and Logging / Where to Log	on		Add sequence number to syslog messages to avoid duplicate suppression.
syslog_split_messages	Reporting and Logging / Where to Log	on		Split messages sent to syslog by lines and to fit into 1024 bytes.
tcp_keepalives_count	Client Connection Defaults / Other Defaults	6		Maximum number of TCP keepalive retransmits.
tcp_keepalives_idle	Client Connection Defaults / Other Defaults	300	s	Time between issuing TCP keepalives.
tcp_keepalives_interval	Client Connection Defaults / Other Defaults	30	s	Time between TCP keepalive retransmits.
tcp_user_timeout	Client Connection Defaults / Other Defaults	0	ms	TCP user timeout.
temp_buffers	Resource Usage / Memory	1024	8kB	Sets the maximum number of temporary buffers used by each session.
temp_file_limit	Resource Usage / Disk	-1	kB	Limits the total size of all temporary files used by each process.
temp_tablespaces	Client Connection Defaults / Statement Behavior			Sets the tablespace(s) to use for temporary tables and sort files.
timezone_abbreviations	Client Connection Defaults / Locale and Formatting	Default		Selects a file of time zone abbreviations.
trace_notify	Developer Options	off		Generates debugging output for LISTEN and NOTIFY.
trace_recovery_messages	Developer Options	log		Enables logging of recovery-related debugging information.
trace_sort	Developer Options	off		Emit information about resource usage in sorting.
track_activities	Statistics / Query and Index Statistics Collector	on		Collects information about executing commands.
track_activity_query_size	Resource Usage / Memory	1024	B	Sets the size reserved for pg_stat_activity.query, in bytes.
track_commit_timestamp	Replication	on		Collects transaction commit time.
track_counts	Statistics / Query and Index Statistics Collector	on		Collects statistics on database activity.
track_functions	Statistics / Query and Index Statistics Collector	none		Collects function-level statistics on database activity.
track_io_timing	Statistics / Query and Index Statistics Collector	on		Collects timing statistics for database I/O activity.
transaction_deferrable	Client Connection Defaults / Statement Behavior	off		Whether to defer a read-only serializable transaction until it can be executed with no possible serialization failures.
transaction_isolation	Client Connection Defaults / Statement Behavior	read committed		Sets the current transaction’s isolation level.
transaction_read_only	Client Connection Defaults / Statement Behavior	off		Sets the current transaction’s read-only status.
transform_null_equals	Version and Platform Compatibility / Other Platforms and Clients	off		Treats “expr=NULL” as “expr IS NULL”.
unix_socket_group	Connections and Authentication / Connection Settings			Sets the owning group of the Unix-domain socket.
unix_socket_permissions	Connections and Authentication / Connection Settings	0777		Sets the access permissions of the Unix-domain socket.
update_process_title	Process Title	on		Updates the process title to show the active SQL command.
vacuum_cleanup_index_scale_factor	Client Connection Defaults / Statement Behavior	0.1		Number of tuple inserts prior to index cleanup as a fraction of reltuples.
vacuum_cost_delay	Resource Usage / Cost-Based Vacuum Delay	0	ms	Vacuum cost delay in milliseconds.
vacuum_cost_limit	Resource Usage / Cost-Based Vacuum Delay	200		Vacuum cost amount available before napping.
vacuum_cost_page_dirty	Resource Usage / Cost-Based Vacuum Delay	20		Vacuum cost for a page dirtied by vacuum.
vacuum_cost_page_hit	Resource Usage / Cost-Based Vacuum Delay	1		Vacuum cost for a page found in the buffer cache.
vacuum_cost_page_miss	Resource Usage / Cost-Based Vacuum Delay	10		Vacuum cost for a page not found in the buffer cache.
vacuum_defer_cleanup_age	Replication / Master Server	0		Number of transactions by which VACUUM and HOT cleanup should be deferred, if any.
vacuum_freeze_min_age	Client Connection Defaults / Statement Behavior	50000000		Minimum age at which VACUUM should freeze a table row.
vacuum_freeze_table_age	Client Connection Defaults / Statement Behavior	150000000		Age at which VACUUM should scan whole table to freeze tuples.
vacuum_multixact_freeze_min_age	Client Connection Defaults / Statement Behavior	5000000		Minimum age at which VACUUM should freeze a MultiXactId in a table row.
vacuum_multixact_freeze_table_age	Client Connection Defaults / Statement Behavior	150000000		Multixact age at which VACUUM should scan whole table to freeze tuples.
wal_block_size	Preset Options	8192		Shows the block size in the write ahead log.
wal_buffers	Write-Ahead Log / Settings	2048	8kB	Sets the number of disk-page buffers in shared memory for WAL.
wal_compression	Write-Ahead Log / Settings	on		Compresses full-page writes written in WAL file.
wal_consistency_checking	Developer Options			Sets the WAL resource managers for which WAL consistency checks are done.
wal_init_zero	Write-Ahead Log / Settings	on		Writes zeroes to new WAL files before first use.
wal_keep_segments	Replication / Sending Servers	16		Sets the number of WAL files held for standby servers.
wal_level	Write-Ahead Log / Settings	replica		Set the level of information written to the WAL.
wal_log_hints	Write-Ahead Log / Settings	on		Writes full pages to WAL when first modified after a checkpoint, even for a non-critical modifications.
wal_receiver_status_interval	Replication / Standby Servers	10	s	Sets the maximum interval between WAL receiver status reports to the sending server.
wal_receiver_timeout	Replication / Standby Servers	60000	ms	Sets the maximum wait time to receive data from the sending server.
wal_recycle	Write-Ahead Log / Settings	on		Recycles WAL files by renaming them.
wal_retrieve_retry_interval	Replication / Standby Servers	5000	ms	Sets the time to wait before retrying to retrieve WAL after a failed attempt.
wal_segment_size	Preset Options	16777216	B	Shows the size of write ahead log segments.
wal_sender_timeout	Replication / Sending Servers	60000	ms	Sets the maximum time to wait for WAL replication.
wal_sync_method	Write-Ahead Log / Settings	fdatasync		Selects the method used for forcing WAL updates to disk.
wal_writer_delay	Write-Ahead Log / Settings	200	ms	Time between WAL flushes performed in the WAL writer.
wal_writer_flush_after	Write-Ahead Log / Settings	128	8kB	Amount of WAL written out by WAL writer that triggers a flush.
work_mem	Resource Usage / Memory	15728	kB	Sets the maximum memory to be used for query workspaces.
xmlbinary	Client Connection Defaults / Statement Behavior	base64		Sets how binary values are to be encoded in XML.
xmloption	Client Connection Defaults / Statement Behavior	content		Sets whether XML data in implicit parsing and serialization operations is to be considered as documents or content fragments.
zero_damaged_pages	Developer Options	off		Continues processing past damaged page headers.

Complete List of Resource Counts

Resource counts prior to running the ingestion load:

resource_type	count
Observation	28643640
Claim	9809811
Procedure	9533768
Encounter	8788209
ExplanationOfBenefit	7795369
DiagnosticReport	2942204
MedicationRequest	2761835
Condition	1755335
Immunization	1656535
CarePlan	1320978
Goal	347835
CareTeam	341226
Practitioner	319783
Organization	319729
AllergyIntolerance	216350
Patient	174388
ImagingStudy	131088
MedicationAdministration	58620
Device	5444
ValueSet	688
CodeSystem	533
Parameters	196
StructureMap	162
Task	155
ConceptMap	94
Questionnaire	68
Contract	66
QuestionnaireResponse	59
ActivityDefinition	44
Location	43
MedicationDispense	40
Bundle	39
PlanDefinition	37
FamilyMemberHistory	32
Medication	32
EvidenceVariable	29
Library	28
Coverage	28
ServiceRequest	28
Group	27
MedicationKnowledge	26
RequestGroup	25
ClaimResponse	25
DeviceRequest	21
Schedule	21
Appointment	21
SubstanceSpecification	18
Person	17
Consent	17
RelatedPerson	17
List	17
AuditEvent	17
NutritionOrder	17
CoverageEligibilityResponse	16
Slot	16
GuidanceResponse	16
Account	15
Communication	15
CommunicationRequest	15
DeviceUseStatement	15
Flag	15
Measure	15
ResearchElementDefinition	15
AdverseEvent	14
ChargeItem	14
MedicationStatement	14
HealthcareService	14
PractitionerRole	13
RiskAssessment	13
AppointmentResponse	13
SupplyRequest	13
Media	13
DetectedIssue	13
Provenance	12
Specimen	11
CompartmentDefinition	11
CoverageEligibilityRequest	11
Composition	11
SupplyDelivery	11
CapabilityStatement	11
ImmunizationEvaluation	10
DeviceDefinition	10
OperationOutcome	10
EventDefinition	10
BiologicallyDerivedProduct	10
Invoice	10
ClinicalImpression	10
MessageDefinition	10
Substance	9
MessageHeader	9
MedicinalProduct	9
MedicinalProductContraindication	8
Endpoint	8
MedicinalProductAuthorization	8
MedicinalProductIndication	8
MedicinalProductInteraction	8
ResearchDefinition	8
SpecimenDefinition	8
SubstanceReferenceInformation	8
Basic	8
ImmunizationRecommendation	7
Binary	7
DocumentManifest	7
ChargeItemDefinition	7
EnrollmentResponse	6
EpisodeOfCare	6
GraphDefinition	6
NamingSystem	6
CatalogEntry	6
DocumentReference	6
ExampleScenario	6
Evidence	6
OrganizationAffiliation	6
EffectEvidenceSynthesis	6
EnrollmentRequest	6
BodyStructure	6
ObservationDefinition	5
PaymentReconciliation	5
MedicinalProductUndesirableEffect	5
MedicinalProductPharmaceutical	5
MedicinalProductPackaged	5
MedicinalProductManufactured	5
MedicinalProductIngredient	5
VisionPrescription	5
ResearchStudy	5
ResearchSubject	5
RiskEvidenceSynthesis	5
Subscription	5
PaymentNotice	5
InsurancePlan	4
VerificationResult	4
DeviceMetric	4
SubstancePolymer	3
SubstanceNucleicAcid	3
TerminologyCapabilities	3
SubstanceProtein	3
SubstanceSourceMaterial	3
ImplementationGuide	3
MeasureReport	2

Edit this page on GitHub

Blog: The Everything Operation

Blog: FHIR Server on Raspberry Pi