MongoDB Input Plugin unexcepted output

[TonyXiaoCui]

mongo v3.2.10
telegraf 1.19

I choose telegraf+prom+grafana to build my montior
I want to collect topstats of collections
when I check top_stat I found unexcepted output
1、one collection with node_type="PRI" but hostname was cfgsrv ip ,not the primary ip
2、cfgsrv has three rs ,so the output also become three
however my telegraf config do not include cfgsrv ips

I learn that the data is from {top:1} but how do it compare its hostname

[lukhas]

Same kind of new unexpected behavior here on 1.19.2, while running on a delayed secondary, I get node_type=PRI, presumably reporting the status of the PRIMARY member of the cluster:

root@late2:~# cat /etc/telegraf/telegraf.d/mongodb.conf
[[inputs.mongodb]]
  servers = ["mongodb://127.0.0.1:27017"]
  gather_perdb_stats = true

Excerpt from rs.config():

		{
			"_id" : 42,
			"host" : "late2:27017",
			"arbiterOnly" : false,
			"buildIndexes" : true,
			"hidden" : false,
			"priority" : 0,
			"tags" : {
				
			},
			"slaveDelay" : NumberLong(86400),
			"votes" : 0
		},

And yet, telegraf sends:

> mongodb,host=late2,hostname=127.0.0.1:27017,node_type=PRI,rs_name=lichess-rs0 active_reads=1i,active_writes=0i,aggregate_command_failed=0i,aggregate_command_total=224884094i,assert_msg=0i,assert_regular=0i,assert_rollovers=0i,assert_user=1085317i,assert_warning=0i,available_reads=127i,available_writes=128i,commands=44658597008i,commands_per_sec=2274i,connections_available=51040i,connections_current=160i,connections_total_created=242705i,count_command_failed=0i,count_command_total=8076090577i,cursor_no_timeout=0i,cursor_no_timeout_count=0i,cursor_pinned=0i,cursor_pinned_count=1i,cursor_timed_out=0i,cursor_timed_out_count=393329319i,cursor_total=15i,cursor_total_count=17213i,delete_command_failed=0i,delete_command_total=209781533i,deletes=419552236i,deletes_per_sec=114i,distinct_command_failed=0i,distinct_command_total=4033627632i,document_deleted=126426616i,document_inserted=1253618366i,document_returned=112488843511i,document_updated=12846861806i,find_and_modify_command_failed=255i,find_and_modify_command_total=1269857197i,find_command_failed=5i,find_command_total=56721750838i,flushes=92229i,flushes_per_sec=0i,flushes_total_time_ns=3758998488000000i,get_more_command_failed=0i,get_more_command_total=25083354195i,getmores=25083354194i,getmores_per_sec=833i,insert_command_failed=0i,insert_command_total=1236361911i,inserts=1256014897i,inserts_per_sec=52i,jumbo_chunks=0i,latency_commands=2753967494006i,latency_commands_count=31062811019i,latency_reads=18614661745404i,latency_reads_count=94139676023i,latency_writes=4737412538863i,latency_writes_count=14643728288i,member_status="PRI",net_in_bytes=4152637i,net_in_bytes_count=88472109634107i,net_out_bytes=4275456i,net_out_bytes_count=98014337610828i,open_connections=160i,operation_scan_and_order=1886718042i,operation_write_conflicts=60967361i,page_faults=6842i,percent_cache_dirty=2.1,percent_cache_used=80,queries=56721750836i,queries_per_sec=2575i,queued_reads=0i,queued_writes=0i,repl_apply_batches_num=343i,repl_apply_batches_total_millis=184i,repl_apply_ops=6722i,repl_buffer_count=0i,repl_buffer_size_bytes=0i,repl_commands=0i,repl_commands_per_sec=0i,repl_deletes=5006i,repl_deletes_per_sec=0i,repl_executor_pool_in_progress_count=0i,repl_executor_queues_network_in_progress=0i,repl_executor_queues_sleepers=7i,repl_executor_unsignaled_events=0i,repl_getmores=0i,repl_getmores_per_sec=0i,repl_inserts=1i,repl_inserts_per_sec=0i,repl_lag=0i,repl_network_bytes=428682i,repl_network_getmores_num=756i,repl_network_getmores_total_millis=9363i,repl_network_ops=2708i,repl_oplog_window_sec=129140i,repl_queries=0i,repl_queries_per_sec=0i,repl_state=1i,repl_updates=1714i,repl_updates_per_sec=0i,resident_megabytes=79325i,state="PRIMARY",storage_freelist_search_bucket_exhausted=0i,storage_freelist_search_requests=0i,storage_freelist_search_scanned=0i,tcmalloc_central_cache_free_bytes=2627170240i,tcmalloc_current_allocated_bytes=79543192920i,tcmalloc_current_total_thread_cache_bytes=88278144i,tcmalloc_heap_size=106448125952i,tcmalloc_max_total_thread_cache_bytes=1073741824i,tcmalloc_pageheap_commit_count=3231612432i,tcmalloc_pageheap_committed_bytes=82928218112i,tcmalloc_pageheap_decommit_count=3188095833i,tcmalloc_pageheap_free_bytes=669548544i,tcmalloc_pageheap_reserve_count=17095i,tcmalloc_pageheap_scavenge_count=3188095833i,tcmalloc_pageheap_total_commit_bytes=5007844158550016i,tcmalloc_pageheap_total_decommit_bytes=5007761230331904i,tcmalloc_pageheap_total_reserve_bytes=106448125952i,tcmalloc_pageheap_unmapped_bytes=23519907840i,tcmalloc_spinlock_total_delay_ns=17454082778400i,tcmalloc_thread_cache_free_bytes=88264576i,tcmalloc_total_free_bytes=2715467200i,tcmalloc_transfer_cache_free_bytes=19328i,total_available=0i,total_created=0i,total_docs_scanned=382088105447i,total_in_use=0i,total_keys_scanned=570086922602i,total_refreshing=0i,total_tickets_reads=128i,total_tickets_writes=128i,ttl_deletes=481761547i,ttl_deletes_per_sec=0i,ttl_passes=155038i,ttl_passes_per_sec=0i,update_command_failed=0i,update_command_total=11927759674i,updates=11928279599i,updates_per_sec=561i,uptime_ns=9354176505000000i,version="4.4.5",vsize_megabytes=103944i,wtcache_app_threads_page_read_count=53111120533i,wtcache_app_threads_page_read_time=2410027540889i,wtcache_app_threads_page_write_count=11155918206i,wtcache_bytes_read_into=1123123212170522i,wtcache_bytes_written_from=253320223993913i,wtcache_current_bytes=80137150987i,wtcache_internal_pages_evicted=6390688874i,wtcache_max_bytes_configured=100166270976i,wtcache_modified_pages_evicted=8560300759i,wtcache_pages_evicted_by_app_thread=0i,wtcache_pages_queued_for_eviction=57869783433i,wtcache_pages_read_into=53141080362i,wtcache_pages_requested_from=1618805424364i,wtcache_pages_written_from=14910353666i,wtcache_server_evicting_pages=0i,wtcache_tracked_dirty_bytes=2076387999i,wtcache_unmodified_pages_evicted=48137684146i,wtcache_worker_thread_evictingpages=57795868572i 1627626282000000000

[arnnow]

Hi,

I have the same kind of behaviour with telegraf 1.19.3.
rollback to 1.19.1 fixed it

Here are the data for a secondary node from influxDB when the rollback was made.

> select connections_current,node_type from mongodb where host = 'mymongohost' and time < '2021-08-25 15:15:00' and time > '2021-08-25 15:03:00'
name: mongodb
time                connections_current node_type
----                ------------------- ---------
1629903810000000000 1200                PRI
1629903840000000000 1204                PRI
1629903870000000000 1208                PRI
1629903900000000000 1208                PRI
1629904410000000000 1135                SEC
1629904440000000000 1138                SEC
1629904470000000000 1139                SEC

mongo is in v4.2.14

[alx75]

We do have the same problem. I skimmed through the code and I wonder if the problem is because in the commit that introduced the new mongo driver (https://github.com/influxdata/telegraf/pull/9493/files) we do not set the RunCmdOptions.ReadPreference and by default it's nil which means primary read (runcmdoptions.go) .

I didn't get the time yet to put something together and test the theory

[arnnow]

I just switch the configuration to the following :

servers = ["mongodb://127.0.0.1:27017/?connect=direct"]

as per noted here : https://pkg.go.dev/go.mongodb.org/mongo-driver/mongo#Connect

Seems to do the trick.
@alx75 same conclusion i guess.

[alx75]

IMO it would be worth putting this info in the mongodb input README.

This is also worth being mentioned in the changelog since this is a breaking change.