1VARNISH-COUNTERS(7) VARNISH-COUNTERS(7)
2
6 varnish-counters - Varnish counter field definitions
7
9 summs – counter - debug
10 stat summ operations
11 Number of times per-thread statistics were summed into the global
13 counters.
14 uptime – counter - info
16 Child process uptime
17 How long the child process has been running.
19 sess_conn – counter - info
21 Sessions accepted
22 Count of sessions successfully accepted
24 sess_drop – counter - info
26 Sessions dropped
27 Count of sessions silently dropped due to lack of worker thread.
29 sess_fail – counter - info
31 Session accept failures
32 Count of failures to accept TCP connection.
34 This counter is the sum of the sess_fail_* counters, which give more
36 detailed information.
37 sess_fail_econnaborted – counter - info
39 Session accept failures: connection aborted
40 Detailed reason for sess_fail: Connection aborted by the client,
42 usually harmless.
43 sess_fail_eintr – counter - info
45 Session accept failures: interrupted system call
46 Detailed reason for sess_fail: The accept() call was interrupted,
48 usually harmless
49 sess_fail_emfile – counter - info
51 Session accept failures: too many open files
52 Detailed reason for sess_fail: No file descriptor was available.
54 Consider raising RLIMIT_NOFILE (see ulimit -n).
55 sess_fail_ebadf – counter - info
57 Session accept failures: bad file descriptor
58 Detailed reason for sess_fail: The listen socket file descriptor was
60 invalid. Should never happen.
61 sess_fail_enomem – counter - info
63 Session accept failures: not enough memory
64 Detailed reason for sess_fail: Most likely insufficient socket buf‐
66 fer memory. Should never happen
67 sess_fail_other – counter - info
69 Session accept failures: other
70 Detailed reason for sess_fail: neither of the above, see SessError
72 log (varnishlog -g raw -i SessError).
73 client_req_400 – counter - info
75 Client requests received, subject to 400 errors
76 400 means we couldn't make sense of the request, it was malformed in
78 some drastic way.
79 client_req_417 – counter - info
81 Client requests received, subject to 417 errors
82 417 means that something went wrong with an Expect: header.
84 client_req – counter - info
86 Good client requests received
87 The count of parseable client requests seen.
89 cache_hit – counter - info
91 Cache hits
92 Count of cache hits. A cache hit indicates that an object has been
94 delivered to a client without fetching it from a backend server.
95 cache_hit_grace – counter - info
97 Cache grace hits
98 Count of cache hits with grace. A cache hit with grace is a cache
100 hit where the object is expired. Note that such hits are also
101 included in the cache_hit counter.
102 cache_hitpass – counter - info
104 Cache hits for pass.
105 Count of hits for pass. A cache hit for pass indicates that Varnish
107 is going to pass the request to the backend and this decision has
108 been cached in it self. This counts how many times the cached deci‐
109 sion is being used.
110 cache_hitmiss – counter - info
112 Cache hits for miss.
113 Count of hits for miss. A cache hit for miss indicates that Varnish
115 is going to proceed as for a cache miss without request coalescing,
116 and this decision has been cached. This counts how many times the
117 cached decision is being used.
118 cache_miss – counter - info
120 Cache misses
121 Count of misses. A cache miss indicates the object was fetched from
123 the backend before delivering it to the client.
124 backend_conn – counter - info
126 Backend conn. success
127 How many backend connections have successfully been established.
129 backend_unhealthy – counter - info
131 Backend conn. not attempted
132 backend_busy – counter - info
134 Backend conn. too many
135 backend_fail – counter - info
137 Backend conn. failures
138 backend_reuse – counter - info
140 Backend conn. reuses
141 Count of backend connection reuses. This counter is increased when‐
143 ever we reuse a recycled connection.
144 backend_recycle – counter - info
146 Backend conn. recycles
147 Count of backend connection recycles. This counter is increased
149 whenever we have a keep-alive connection that is put back into the
150 pool of connections. It has not yet been used, but it might be,
151 unless the backend closes it.
152 backend_retry – counter - info
154 Backend conn. retry
155 fetch_head – counter - info
157 Fetch no body (HEAD)
158 beresp with no body because the request is HEAD.
160 fetch_length – counter - info
162 Fetch with Length
163 beresp.body with Content-Length.
165 fetch_chunked – counter - info
167 Fetch chunked
168 beresp.body with Chunked.
170 fetch_eof – counter - info
172 Fetch EOF
173 beresp.body with EOF.
175 fetch_bad – counter - info
177 Fetch bad T-E
178 beresp.body length/fetch could not be determined.
180 fetch_none – counter - info
182 Fetch no body
183 beresp.body empty
185 fetch_1xx – counter - info
187 Fetch no body (1xx)
188 beresp with no body because of 1XX response.
190 fetch_204 – counter - info
192 Fetch no body (204)
193 beresp with no body because of 204 response.
195 fetch_304 – counter - info
197 Fetch no body (304)
198 beresp with no body because of 304 response.
200 fetch_failed – counter - info
202 Fetch failed (all causes)
203 beresp fetch failed.
205 fetch_no_thread – counter - info
207 Fetch failed (no thread)
208 beresp fetch failed, no thread available.
210 pools – gauge - info
212 Number of thread pools
213 Number of thread pools. See also parameter thread_pools. NB:
215 Presently pools cannot be removed once created.
216 threads – gauge - info
218 Total number of threads
219 Number of threads in all pools. See also parameters thread_pools,
221 thread_pool_min and thread_pool_max.
222 threads_limited – counter - info
224 Threads hit max
225 Number of times more threads were needed, but limit was reached in a
227 thread pool. See also parameter thread_pool_max.
228 threads_created – counter - info
230 Threads created
231 Total number of threads created in all pools.
233 threads_destroyed – counter - info
235 Threads destroyed
236 Total number of threads destroyed in all pools.
238 threads_failed – counter - info
240 Thread creation failed
241 Number of times creating a thread failed. See VSL::Debug for diag‐
243 nostics. See also parameter thread_fail_delay.
244 thread_queue_len – gauge - info
246 Length of session queue
247 Length of session queue waiting for threads. NB: Only updates once
249 per second. See also parameter thread_queue_limit.
250 busy_sleep – counter - info
252 Number of requests sent to sleep on busy objhdr
253 Number of requests sent to sleep without a worker thread because
255 they found a busy object.
256 busy_wakeup – counter - info
258 Number of requests woken after sleep on busy objhdr
259 Number of requests taken off the busy object sleep list and resched‐
261 uled.
262 busy_killed – counter - info
264 Number of requests killed after sleep on busy objhdr
265 Number of requests killed from the busy object sleep list due to
267 lack of resources.
268 sess_queued – counter - info
270 Sessions queued for thread
271 Number of times session was queued waiting for a thread. See also
273 parameter thread_queue_limit.
274 sess_dropped – counter - info
276 Sessions dropped for thread
277 Number of times an HTTP/1 session was dropped because the queue was
279 too long already. See also parameter thread_queue_limit.
280 req_dropped – counter - info
282 Requests dropped
283 Number of times an HTTP/2 stream was refused because the queue was
285 too long already. See also parameter thread_queue_limit.
286 n_object – gauge - info
288 object structs made
289 Approximate number of HTTP objects (headers + body, if present) in
291 the cache.
292 n_vampireobject – gauge - diag
294 unresurrected objects
295 Number of unresurrected objects
297 n_objectcore – gauge - info
299 objectcore structs made
300 Approximate number of object metadata elements in the cache. Each
302 object needs an objectcore, extra objectcores are for hit-for-miss,
303 hit-for-pass and busy objects.
304 n_objecthead – gauge - info
306 objecthead structs made
307 Approximate number of different hash entries in the cache.
309 n_backend – gauge - info
311 Number of backends
312 Number of backends known to us.
314 n_expired – counter - info
316 Number of expired objects
317 Number of objects that expired from cache because of old age.
319 n_lru_nuked – counter - info
321 Number of LRU nuked objects
322 How many objects have been forcefully evicted from storage to make
324 room for a new object.
325 n_lru_moved – counter - diag
327 Number of LRU moved objects
328 Number of move operations done on the LRU list.
330 n_lru_limited – counter - info
332 Reached nuke_limit
333 Number of times more storage space were needed, but limit was
335 reached in a nuke_limit. See also parameter nuke_limit.
336 losthdr – counter - info
338 HTTP header overflows
339 s_sess – counter - info
341 Total sessions seen
342 n_pipe – gauge - info
344 Number of ongoing pipe sessions
345 pipe_limited – counter - info
347 Pipes hit pipe_sess_max
348 Number of times more pipes were needed, but the limit was reached.
350 See also parameter pipe_sess_max.
351 s_pipe – counter - info
353 Total pipe sessions seen
354 s_pass – counter - info
356 Total pass-ed requests seen
357 s_fetch – counter - info
359 Total backend fetches initiated
360 s_synth – counter - info
362 Total synthetic responses made
363 s_req_hdrbytes – counter - info
365 Request header bytes
366 Total request header bytes received
368 s_req_bodybytes – counter - info
370 Request body bytes
371 Total request body bytes received
373 s_resp_hdrbytes – counter - info
375 Response header bytes
376 Total response header bytes transmitted
378 s_resp_bodybytes – counter - info
380 Response body bytes
381 Total response body bytes transmitted
383 s_pipe_hdrbytes – counter - info
385 Pipe request header bytes
386 Total request bytes received for piped sessions
388 s_pipe_in – counter - info
390 Piped bytes from client
391 Total number of bytes forwarded from clients in pipe sessions
393 s_pipe_out – counter - info
395 Piped bytes to client
396 Total number of bytes forwarded to clients in pipe sessions
398 sess_closed – counter - info
400 Session Closed
401 sess_closed_err – counter - info
403 Session Closed with error
404 Total number of sessions closed with errors. See sc_* diag counters
406 for detailed breakdown
407 sess_readahead – counter - info
409 Session Read Ahead
410 sess_herd – counter - diag
412 Session herd
413 Number of times the timeout_linger triggered
415 sc_rem_close – counter - diag
417 Session OK REM_CLOSE
418 Number of session closes with REM_CLOSE (Client Closed)
420 sc_req_close – counter - diag
422 Session OK REQ_CLOSE
423 Number of session closes with REQ_CLOSE (Client requested close)
425 sc_req_http10 – counter - diag
427 Session Err REQ_HTTP10
428 Number of session closes with Error REQ_HTTP10 (Proto < HTTP/1.1)
430 sc_rx_bad – counter - diag
432 Session Err RX_BAD
433 Number of session closes with Error RX_BAD (Received bad req/resp)
435 sc_rx_body – counter - diag
437 Session Err RX_BODY
438 Number of session closes with Error RX_BODY (Failure receiving
440 req.body)
441 sc_rx_junk – counter - diag
443 Session Err RX_JUNK
444 Number of session closes with Error RX_JUNK (Received junk data)
446 sc_rx_overflow – counter - diag
448 Session Err RX_OVERFLOW
449 Number of session closes with Error RX_OVERFLOW (Received buffer
451 overflow)
452 sc_rx_timeout – counter - diag
454 Session Err RX_TIMEOUT
455 Number of session closes with Error RX_TIMEOUT (Receive timeout)
457 sc_tx_pipe – counter - diag
459 Session OK TX_PIPE
460 Number of session closes with TX_PIPE (Piped transaction)
462 sc_tx_error – counter - diag
464 Session Err TX_ERROR
465 Number of session closes with Error TX_ERROR (Error transaction)
467 sc_tx_eof – counter - diag
469 Session OK TX_EOF
470 Number of session closes with TX_EOF (EOF transmission)
472 sc_resp_close – counter - diag
474 Session OK RESP_CLOSE
475 Number of session closes with RESP_CLOSE (Backend/VCL requested
477 close)
478 sc_overload – counter - diag
480 Session Err OVERLOAD
481 Number of session closes with Error OVERLOAD (Out of some resource)
483 sc_pipe_overflow – counter - diag
485 Session Err PIPE_OVERFLOW
486 Number of session closes with Error PIPE_OVERFLOW (Session pipe
488 overflow)
489 sc_range_short – counter - diag
491 Session Err RANGE_SHORT
492 Number of session closes with Error RANGE_SHORT (Insufficient data
494 for range)
495 sc_req_http20 – counter - diag
497 Session Err REQ_HTTP20
498 Number of session closes with Error REQ_HTTP20 (HTTP2 not accepted)
500 sc_vcl_failure – counter - diag
502 Session Err VCL_FAILURE
503 Number of session closes with Error VCL_FAILURE (VCL failure)
505 client_resp_500 – counter - diag
507 Delivery failed due to insufficient workspace.
508 Number of times we failed a response due to running out of workspace
510 memory during delivery.
511 ws_backend_overflow – counter - diag
513 workspace_backend overflows
514 Number of times we ran out of space in workspace_backend.
516 ws_client_overflow – counter - diag
518 workspace_client overflows
519 Number of times we ran out of space in workspace_client.
521 ws_thread_overflow – counter - diag
523 workspace_thread overflows
524 Number of times we ran out of space in workspace_thread.
526 ws_session_overflow – counter - diag
528 workspace_session overflows
529 Number of times we ran out of space in workspace_session.
531 shm_records – counter - diag
533 SHM records
534 shm_writes – counter - diag
536 SHM writes
537 shm_flushes – counter - diag
539 SHM flushes due to overflow
540 shm_cont – counter - diag
542 SHM MTX contention
543 shm_cycles – counter - diag
545 SHM cycles through buffer
546 backend_req – counter - info
548 Backend requests made
549 n_vcl – gauge - info
551 Number of loaded VCLs in total
552 n_vcl_avail – gauge - diag
554 Number of VCLs available
555 n_vcl_discard – gauge - diag
557 Number of discarded VCLs
558 vcl_fail – counter - info
560 VCL failures
561 Count of failures which prevented VCL from completing.
563 bans – gauge - info
565 Count of bans
566 Number of all bans in system, including bans superseded by newer
568 bans and bans already checked by the ban-lurker.
569 bans_completed – gauge - diag
571 Number of bans marked 'completed'
572 Number of bans which are no longer active, either because they got
574 checked by the ban-lurker or superseded by newer identical bans.
575 bans_obj – gauge - diag
577 Number of bans using obj.*
578 Number of bans which use obj.* variables. These bans can possibly
580 be washed by the ban-lurker.
581 bans_req – gauge - diag
583 Number of bans using req.*
584 Number of bans which use req.* variables. These bans can not be
586 washed by the ban-lurker.
587 bans_added – counter - diag
589 Bans added
590 Counter of bans added to ban list.
592 bans_deleted – counter - diag
594 Bans deleted
595 Counter of bans deleted from ban list.
597 bans_tested – counter - diag
599 Bans tested against objects (lookup)
600 Count of how many bans and objects have been tested against each
602 other during hash lookup.
603 bans_obj_killed – counter - diag
605 Objects killed by bans (lookup)
606 Number of objects killed by bans during object lookup.
608 bans_lurker_tested – counter - diag
610 Bans tested against objects (lurker)
611 Count of how many bans and objects have been tested against each
613 other by the ban-lurker.
614 bans_tests_tested – counter - diag
616 Ban tests tested against objects (lookup)
617 Count of how many tests and objects have been tested against each
619 other during lookup. 'ban req.url == foo && req.http.host == bar'
620 counts as one in 'bans_tested' and as two in 'bans_tests_tested'
621 bans_lurker_tests_tested – counter - diag
623 Ban tests tested against objects (lurker)
624 Count of how many tests and objects have been tested against each
626 other by the ban-lurker. 'ban req.url == foo && req.http.host ==
627 bar' counts as one in 'bans_tested' and as two in
628 'bans_tests_tested'
629 bans_lurker_obj_killed – counter - diag
631 Objects killed by bans (lurker)
632 Number of objects killed by the ban-lurker.
634 bans_lurker_obj_killed_cutoff – counter - diag
636 Objects killed by bans for cutoff (lurker)
637 Number of objects killed by the ban-lurker to keep the number of
639 bans below ban_cutoff.
640 bans_dups – counter - diag
642 Bans superseded by other bans
643 Count of bans replaced by later identical bans.
645 bans_lurker_contention – counter - diag
647 Lurker gave way for lookup
648 Number of times the ban-lurker had to wait for lookups.
650 bans_persisted_bytes – gauge - diag
652 Bytes used by the persisted ban lists
653 Number of bytes used by the persisted ban lists.
655 bans_persisted_fragmentation – gauge - diag
657 Extra bytes in persisted ban lists due to fragmentation
658 Number of extra bytes accumulated through dropped and completed bans
660 in the persistent ban lists.
661 n_purges – counter - info
663 Number of purge operations executed
664 n_obj_purged – counter - info
666 Number of purged objects
667 exp_mailed – counter - diag
669 Number of objects mailed to expiry thread
670 Number of objects mailed to expiry thread for handling.
672 exp_received – counter - diag
674 Number of objects received by expiry thread
675 Number of objects received by expiry thread for handling.
677 hcb_nolock – counter - debug
679 HCB Lookups without lock
680 hcb_lock – counter - debug
682 HCB Lookups with lock
683 hcb_insert – counter - debug
685 HCB Inserts
686 esi_errors – counter - diag
688 ESI parse errors (unlock)
689 esi_warnings – counter - diag
691 ESI parse warnings (unlock)
692 vmods – gauge - info
694 Loaded VMODs
695 n_gzip – counter - info
697 Gzip operations
698 n_gunzip – counter - info
700 Gunzip operations
701 n_test_gunzip – counter - info
703 Test gunzip operations
704 Those operations occur when Varnish receives a compressed object
706 from a backend. They are done to verify the gzip stream while it's
707 inserted in storage.
708
710 uptime – counter - info
711 Management process uptime
712 Uptime in seconds of the management process
714 child_start – counter - diag
716 Child process started
717 Number of times the child process has been started
719 child_exit – counter - diag
721 Child process normal exit
722 Number of times the child process has been cleanly stopped
724 child_stop – counter - diag
726 Child process unexpected exit
727 Number of times the child process has exited with an unexpected
729 return code
730 child_died – counter - diag
732 Child process died (signal)
733 Number of times the child process has died due to signals
735 child_dump – counter - diag
737 Child process core dumped
738 Number of times the child process has produced core dumps
740 child_panic – counter - diag
742 Child process panic
743 Number of times the management process has caught a child panic
745
747 live – gauge - debug
748 In use
749 pool – gauge - debug
751 In Pool
752 sz_wanted – gauge - debug
754 Size requested
755 sz_actual – gauge - debug
757 Size allocated
758 allocs – counter - debug
760 Allocations
761 frees – counter - debug
763 Frees
764 recycle – counter - debug
766 Recycled from pool
767 timeout – counter - debug
769 Timed out from pool
770 toosmall – counter - debug
772 Too small to recycle
773 surplus – counter - debug
775 Too many for pool
776 randry – counter - debug
778 Pool ran dry
779
781 c_req – counter - info
782 Allocator requests
783 Number of times the storage has been asked to provide a storage seg‐
785 ment.
786 c_fail – counter - info
788 Allocator failures
789 Number of times the storage has failed to provide a storage segment.
791 c_bytes – counter - info
793 Bytes allocated
794 Number of total bytes allocated by this storage.
796 c_freed – counter - info
798 Bytes freed
799 Number of total bytes returned to this storage.
801 g_alloc – gauge - info
803 Allocations outstanding
804 Number of storage allocations outstanding.
806 g_bytes – gauge - info
808 Bytes outstanding
809 Number of bytes allocated from the storage.
811 g_space – gauge - info
813 Bytes available
814 Number of bytes left in the storage.
816
818 c_req – counter - info
819 Allocator requests
820 Number of times the storage has been asked to provide a storage seg‐
822 ment.
823 c_fail – counter - info
825 Allocator failures
826 Number of times the storage has failed to provide a storage segment.
828 c_bytes – counter - info
830 Bytes allocated
831 Number of total bytes allocated by this storage.
833 c_freed – counter - info
835 Bytes freed
836 Number of total bytes returned to this storage.
838 g_alloc – gauge - info
840 Allocations outstanding
841 Number of storage allocations outstanding.
843 g_bytes – gauge - info
845 Bytes outstanding
846 Number of bytes allocated from the storage.
848 g_space – gauge - info
850 Bytes available
851 Number of bytes left in the storage.
853
855 c_req – counter - info
856 Allocator requests
857 Number of times the storage has been asked to provide a storage seg‐
859 ment.
860 c_fail – counter - info
862 Allocator failures
863 Number of times the storage has failed to provide a storage segment.
865 c_bytes – counter - info
867 Bytes allocated
868 Number of total bytes allocated by this storage.
870 c_freed – counter - info
872 Bytes freed
873 Number of total bytes returned to this storage.
875 g_alloc – gauge - info
877 Allocations outstanding
878 Number of storage allocations outstanding.
880 g_bytes – gauge - info
882 Bytes outstanding
883 Number of bytes allocated from the storage.
885 g_space – gauge - info
887 Bytes available
888 Number of bytes left in the storage.
890 g_smf – gauge - info
892 N struct smf
893 g_smf_frag – gauge - info
895 N small free smf
896 g_smf_large – gauge - info
898 N large free smf
899
901 happy – bitmap - info
902 Happy health probes
903 bereq_hdrbytes – counter - info
905 Request header bytes
906 Total backend request header bytes sent
908 bereq_bodybytes – counter - info
910 Request body bytes
911 Total backend request body bytes sent
913 beresp_hdrbytes – counter - info
915 Response header bytes
916 Total backend response header bytes received
918 beresp_bodybytes – counter - info
920 Response body bytes
921 Total backend response body bytes received
923 pipe_hdrbytes – counter - info
925 Pipe request header bytes
926 Total request bytes sent for piped sessions
928 pipe_out – counter - info
930 Piped bytes to backend
931 Total number of bytes forwarded to backend in pipe sessions
933 pipe_in – counter - info
935 Piped bytes from backend
936 Total number of bytes forwarded from backend in pipe sessions
938 conn – gauge - info
940 Concurrent connections to backend
941 req – counter - info
943 Backend requests sent
944 unhealthy – counter - info
946 Fetches not attempted due to backend being unhealthy
947 busy – counter - info
949 Fetches not attempted due to backend being busy
950 Number of times the max_connections limit was reached
952 fail – counter - info
954 Connections failed
955 Counter of failed opens. Detailed reasons are given in the fail_*
957 counters (DIAG level) and in the log under the FetchError tag.
958 This counter is the sum of all detailed fail_* counters.
960 All fail_* counters may be slightly inaccurate for efficiency.
962 fail_eacces – counter - diag
964 Connections failed with EACCES or EPERM
965 fail_eaddrnotavail – counter - diag
967 Connections failed with EADDRNOTAVAIL
968 fail_econnrefused – counter - diag
970 Connections failed with ECONNREFUSED
971 fail_enetunreach – counter - diag
973 Connections failed with ENETUNREACH
974 fail_etimedout – counter - diag
976 Connections failed ETIMEDOUT
977 fail_other – counter - diag
979 Connections failed for other reason
980 helddown – counter - diag
982 Connection opens not attempted
983 Connections not attempted during the backend_local_error_holddown or
985 backend_remote_error_holddown interval after a fundamental connec‐
986 tion issue.
987
989 Counters which track the activity in the different classes of
990 mutex-locks.
991 The counts may be slightly wrong if there are more than one lock
993 instantiated in each class (ie: .creat > 1)
994 creat – counter - debug
996 Created locks
997 destroy – counter - debug
999 Destroyed locks
1000 locks – counter - debug
1002 Lock Operations
1003 dbg_busy – counter - debug
1005 Contended lock operations
1006 If the lck debug bit is set: Lock operations which returned EBUSY on
1008 the first locking attempt.
1009 If the lck debug bit is unset, this counter will never be incre‐
1011 mented even if lock operations are contended.
1012 dbg_try_fail – counter - debug
1014 Contended trylock operations
1015 If the lck debug bit is set: Trylock operations which returned
1017 EBUSY.
1018 If the lck debug bit is unset, this counter will never be incre‐
1020 mented even if lock operations are contended.
1021
1023 This man page was written by Lasse Karstensen, using content from
1024 vsc2rst written by Tollef Fog Heen.
1025 VARNISH-COUNTERS(7)