All logs are on in the SG config file, including debug logs. This is all I get after restarting the service:
2019-10-11T18:42:54.389+02:00 ==== Couchbase Sync Gateway/2.6.0(127;b4c828d) CE ====
2019-10-11T18:42:54.390+02:00 [INF] Logging: Console to stderr
2019-10-11T18:42:54.390+02:00 [INF] Logging: Files to /home/sync_gateway/logs
2019-10-11T18:42:54.390+02:00 [INF] Logging: Console level: info
2019-10-11T18:42:54.390+02:00 [INF] Logging: Console keys: [HTTP]
2019-10-11T18:42:54.390+02:00 [INF] Logging: Redaction level: none
2019-10-11T18:42:54.390+02:00 [INF] requestedSoftFDLimit >= currentHardFdLimit (250000 >= 65535) capping at 65535
2019-10-11T18:42:54.390+02:00 [INF] Configured process to allow 65535 open file descriptors
2019-10-11T18:42:54.390+02:00 [INF] Logging stats with frequency: 1m0s
2019-10-11T18:42:54.390+02:00 [INF] Opening db /<my_database> as bucket "test_bucket", pool "default", server <couchbase://192.168.178.72,192.168.178.80>
2019-10-11T18:42:54.390+02:00 [WRN] Using deprecated config option: "channel_cache_max_length". Use "cache.channel_cache.max_length" instead. -- rest.(*ServerContext)._getOrAddDatabaseFromConfig() at server_context.go:384
2019-10-11T18:42:54.391+02:00 [INF] GoCBCustomSGTranscoder Opening Couchbase database test_bucket on <couchbase://192.168.178.72,192.168.178.80> as user "sync_gateway_test"
2019-10-11T18:42:54.392+02:00 [INF] Auth: Attempting credential authentication couchbase://192.168.178.72,192.168.178.80?http_idle_conn_timeout=90000&http_max_idle_conns=64000&http_max_idle_conns_per_host=256&n1ql_timeout=75000
2019-10-11T18:42:54.399+02:00 [INF] Successfully opened bucket test_bucket
2019-10-11T18:42:54.410+02:00 [INF] Set query timeouts for bucket test_bucket to cluster:1m15s, bucket:1m15s
2019-10-11T18:42:54.410+02:00 [INF] Initializing indexes with numReplicas: 0...
2019-10-11T18:42:54.723+02:00 [INF] Verifying index availability for bucket test_bucket...
2019-10-11T18:42:54.736+02:00 [INF] Indexes ready for bucket test_bucket.
2019-10-11T18:42:54.736+02:00 [INF] delta_sync enabled=false with rev_max_age_seconds=86400 for database lister_production_database
2019-10-11T18:42:54.737+02:00 [INF] Created background task: "CleanAgedItems" with interval 1m0s
2019-10-11T18:42:54.737+02:00 [INF] Cache: Initializing changes cache with options {ChannelCacheOptions:{ChannelCacheMinLength:50 ChannelCacheMaxLength:1000 ChannelCacheAge:1m0s MaxNumChannels:50000 CompactHighWatermarkPercent:80 CompactLowWatermarkPercent:60} CachePendingSeqMaxWait:5s CachePendingSeqMaxNum:10000 CacheSkippedSeqMaxWait:1h0m0s}
2019-10-11T18:42:54.737+02:00 [INF] Initializing changes cache for database lister_production_database
2019-10-11T18:42:54.737+02:00 [INF] Created background task: "InsertPendingEntries" with interval 2.5s
2019-10-11T18:42:54.737+02:00 [INF] Created background task: "CleanSkippedSequenceQueue" with interval 30m0s
2019-10-11T18:42:54.737+02:00 [INF] DCP: Starting mutation feed on bucket test_bucket due to either channel cache mode or doc tracking (auto-import/bucketshadow)
2019-10-11T18:42:54.737+02:00 [INF] DCP: Using DCP feed for bucket: "test_bucket" (based on feed_type specified in config file)
2019-10-11T18:42:54.738+02:00 [INF] DCP: Using DCP Logging Receiver.
2019-10-11T18:42:54.777+02:00 [INF] DCP: Previously persisted backfill sequences found - will resume
2019-10-11T18:42:54.777+02:00 [INF] DCP: Restarting vb 0 using metadata sequence 0 (backfill 0 not in [0-0])
2019-10-11T18:42:54.777+02:00 [INF] DCP: Restarting vb 1 using metadata sequence 0 (backfill 0 not in [0-0])
2019-10-11T18:42:54.778+02:00 [INF] DCP: Restarting vb 2 using metadata sequence 0 (backfill 0 not in [0-0])
2019-10-11T18:42:54.778+02:00 [INF] DCP: Restarting vb 3 using metadata sequence 0 (backfill 0 not in [0-0])
...
2019-10-11T18:42:55.189+02:00 [INF] DCP: Restarting vb 1022 using metadata sequence 0 (backfill 0 not in [0-0])
2019-10-11T18:42:55.189+02:00 [INF] DCP: Restarting vb 1023 using metadata sequence 0 (backfill 0 not in [0-0])
2019-10-11T18:42:55.204+02:00 [INF] Using metadata purge interval of 3.00 days for tombstone compaction.
2019-10-11T18:42:55.204+02:00 [INF] Created background task: "Compact" with interval 24h0m0s
2019-10-11T18:42:55.211+02:00 [INF] Reset guest user to config
2019-10-11T18:42:55.211+02:00 [INF] Starting admin server on :4985
2019-10-11T18:42:55.218+02:00 [INF] Starting server on :4984 ...
2019-10-11T18:42:55.228+02:00 [INF] CBGoUtilsLogger: Using plain authentication for user <ud>sync_gateway_test</ud>
2019-10-11T18:42:55.228+02:00 [INF] CBGoUtilsLogger: Using plain authentication for user <ud>sync_gateway_test</ud>
2019-10-11T18:42:55.257+02:00 [INF] DCP: Backfill in progress: 7% (1 / 14)
url is: URI syncGatewayURI = new URI("ws://192.168.178.68:4984/<my_database>");
There is 3 seperate machines for loadbalancer, SG, Couchbase Server nodes. Hence 3 different local IPs.