Skip to content

Latest commit

 

History

History
186 lines (138 loc) · 9.02 KB

changelog.md

File metadata and controls

186 lines (138 loc) · 9.02 KB
Raw

Version 20231201

  • Minimum requirements for all software (elasticsearch, logstash, kibana) is 7.17

    • for many reasons, but 1 of many great reasons is Ingest pipeline => 7.11 supports the set processor function "copy_from"

    • tag_with_exception_message for ruby filter block

      • requires version => 7.17
      • requires version => 8.x

    • ingest pipelines finally performant in => 8.8

    • logstash more performant in 8.x

    • many many other reasons

  • [@metadata][corelight_env_vars][disable_legacy_md5_fingerprint_hash] if set to true will disable legacy md5 fingerprint hashes else will call 8811-corelight-ecs-network_fingerprints-enrich-legacy_md5-filter.conf and add an array w/ md5 to the existing sha1

  • changed values for labels.corelight.ecs_method. values are now logstash_pipeline or ingest_pipeline previously were logstash or ingest_pipeline

  • added labels.corelight.ecs_pipeline_method

  • replaced text analyzer with es_stk_analyzer

  • removed stats index, replaced with metrics. event.category is diagnostics same as the other metrics (previously was 'stats') these were the logs:

    • capture_loss
    • corelight_cloud_stats
    • corelight_overall_capture_loss
    • corelight_profiling
    • corelight_weird_stats
    • namecache
    • packet_filter
    • reporter
    • stats
    • suricata_stats
    • weird_stats

  • removed netcontrol and iam index, replaced with system. these were the logs:

    • netcontrol
    • netcontrol_drop
    • netcontrol_shunt
    • corelight_audit_log
    • audit
    • auditlog

  • changed suricata_corelight dataset back to suricata_corelight from suricata

  • parse-failures is now parse_failures & parse-failure is parse_failure

  • change intel log event.category from intrustion_detection to threat

  • changed VAR_$Corelight_LS_Index_Strategy to VAR_CORELIGHT_INDEX_STRATEGY

  • logs that do not explicitly have a pipeline/parser (new or unknown) they get set to event.category tbd (previously used either temporary or event.dataset). not use event.dataset as do not want to autocreate datastreams that would live outside ILM patterns and other control/settings and permissions. can still search the logs as is despite the name in the index pattern. event.dataset still gets set.

  • you can set [@metadata][custom_temporary_metadata_index_name_namespace] or [data_stream][namespace] to set a desired namespace for multiple tenants/clients data

    • if [@metadata][custom_temporary_metadata_index_name_namespace] is set then it overrides if [data_stream][namespace] already exists

  • for custom mappings,settings,ilm,etc.. use

    • aliases

      • corelight-main_logs-aliases@custom
      • corelight-metrics_and_stats-aliases@custom
      • corelight-parse_failures-aliases@custom

    • base settings

      • corelight-main_logs-base-settings@custom
      • corelight-metrics_and_stats-base-settings@custom
      • corelight-parse_failures-base-settings@custom

    • ilm settings

      • corelight-main_logs-ilm-settings@custom
      • corelight-metrics_and_stats-ilm-settings@custom
      • corelight-parse_failures-ilm-settings@custom

    • mappings

      • corelight-main_logs-mappings@custom
      • corelight-metrics_and_stats-mappings@custom
      • corelight-parse_failures-mappings@custom

  • ingest pipelines custom additions user controlled
    if the pipelines do not exist, it just ignores them. so in other words, is optional for user

  • names are

    • before starting corelight ingest pipelines use 'corelight-ecs-main-pipeline@custom'
    • after completing corelight ingest pipelines use 'corelight-ecs-postprocess-final-main-pipeline@custom'

  • date fields that are user/software/attacker controlled (smtp.date for example) or commonly could be invalid (like for x509 cert dates before/after epoch) are always kept now instead of removed. if valid date timestamp then it is copied as such to a new field

  • logstash moved 0101-corelight-ecs-user_defined-set_indexing_strategy-filter.conf into 3100-corelight-ecs-common-set_index_prefix_and_suffix-filter.conf ( no longer need to edit manually anyways)

  • installer script, use last run

    • uses last run, allows directory to be specified
    • this can also be used to just use the script to upload things you generated elsewhere or if already have the files generated

  • additional changelog

    • logstash network fingerprints have md5 and sha1, sha1 will be used for future so that ingest pipelines will be able to support to. and ingest pipelines do not have a builtin callable md5 from script processor. only fingerprint processor that unfortunately has a bug of copying a null byte that makes the hash unusuable with other things (like logstash in this case)

    • single quote invalid IPs in event.tag (previously was not quoted)

    • IPs beginning with "0" (ie: 0.0.0.0/8) changed

      • ip_type from "reserved_as_a_source_address_only" to "reserved_local_this_network"
      • ip_rfc from "RFC_1700" to "RFC_1122-3.2.1.3"

    • split out RFC info for all RFC6890, RFC2544, RFC5737 that were previously tagged together

    • 255.255.255.255 changed from RFC_8190 to RFC_919

    • added rfc for 240.0.0.0/4 RFC5735, no longer grouped under multicast

    • carrier grade nat 100.64.0.0-100.127.255.255 fixed type label

    • 0.0.0.0 set to type reserved_any_address

    • switched ingest pipeline's main_pipeline to call pipeline name w/ the event.dataset

    • changed the "genera"/"common" corelight_genenral_pipeline to corelight_common_pipeline

    • renamed all pipelines (to match similar naming in logstash)

      • corelight_ to corelight-ecs-
      • _pipeline to -pipeline

    • renamed pipelines (to match similar naming in logstash)

      • metrics_general to common-metrics
      • netcontrol_general to common-netcontrol
      • system_related_info_general to common-system
      • stats_general to common-stats

    • extended time fields for pulling out x509 and smtp date fields

    • ECS email field set is GA. previously these were always copied. However, they are now renamed instead. as follows. Alias field mappings are provided for backwards compatibility.

      • smtp.cc > email.cc.address
      • smtp.from > email.from.address
      • smtp.mailfrom > email.sender.address
      • smtp.msg_id > email.message_id
      • smtp.reply_to > email.reply_to.address
      • smtp.subject > email.subject
      • smtp.to > email.to.address
      • smtp.date > smtp.date_non_formatted_date (always kept)
      • smtp.date_non_formatted_date COPIED to email.origination_timestamp (if valid date timestamp)
      • smtp.subject_has_non_ascii > email.subject_has_non_ascii
      • smtp.subject_length > email.subject_length

  • for corelight_weird_stats.log and weird_stats.log

    • changed name from labels.weird_name to name
      • removed config in ingest pipeline, was not in logstash pipeline
    • changed num_seen from labels.weird_number_seen to num_seen
      • removed config in ingest pipeline, was not in logstash pipeline

  • for conn_doctor.log

    • changed bad from conn_doctor.bad to bad
      • removed config in logstash pipeline, was not in ingest pipeline
    • changed check from conn_doctor.check to check
      • removed config in logstash pipeline, was not in ingest pipeline
    • changed percent from conn_doctor.percent to percent
      • removed config in logstash pipeline, was not in ingest pipeline

  • removed labels.dns.query_length from mappings, which was just an alias to dns.question.name_length

  • spcap fields for conn.log

    • ingest pipeline:
      • added varations for underscore fields (ie: spcap_rule, spcap_trigger, spcap_url)
      • spcap_rule from labels.corelight.spcap.rule to conn.spcap_rule
      • spcap_trigger from labels.corelight.spcap_trigger to conn.spcap_trigger
      • spcap_url from labels.corelight.spcap_url to conn.spcap_url
    • logstash pipeline:
      • spcap_rule from labels.corelight.spcap.rule to conn.spcap_rule
      • spcap_trigger from labels.corelight.spcap.trigger to conn.spcap_trigger
      • spcap_url from labels.corelight.spcap.url to conn.spcap_url

  • removed labels.etl.elasticsearch_index_* fields

    • labels.etl.elasticsearch_index_name_prefix
    • labels.etl.elasticsearch_index_name_suffix
    • labels.etl.elasticsearch_index_strategy

  • suricata_corelight log

    • append the value from suricata.alert.action to event.type. previously alert.action was renamed to event.type

  • for metric and system logs, perform '.' fields to '_' for ingest pipeline the same as is done in logstash pipelines

  • http log

    • logstash pipeline
      • username from username to url.username & copy to source.user.name
    • ingest pipeline
      • username from source.user.name to url.username & copy to source.user.name
      • password from source.user.password to url.password & copy to source.user.password