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_fail – counter - info
26 Session accept failures
27 Count of failures to accept TCP connection.
29 This counter is the sum of the sess_fail_* counters, which give more
31 detailed information.
32 sess_fail_econnaborted – counter - info
34 Session accept failures: connection aborted
35 Detailed reason for sess_fail: Connection aborted by the client,
37 usually harmless.
38 sess_fail_eintr – counter - info
40 Session accept failures: interrupted system call
41 Detailed reason for sess_fail: The accept() call was interrupted,
43 usually harmless
44 sess_fail_emfile – counter - info
46 Session accept failures: too many open files
47 Detailed reason for sess_fail: No file descriptor was available.
49 Consider raising RLIMIT_NOFILE (see ulimit -n).
50 sess_fail_ebadf – counter - info
52 Session accept failures: bad file descriptor
53 Detailed reason for sess_fail: The listen socket file descriptor was
55 invalid. Should never happen.
56 sess_fail_enomem – counter - info
58 Session accept failures: not enough memory
59 Detailed reason for sess_fail: Most likely insufficient socket buf‐
61 fer memory. Should never happen
62 sess_fail_other – counter - info
64 Session accept failures: other
65 Detailed reason for sess_fail: neither of the above, see SessError
67 log (varnishlog -g raw -i SessError).
68 client_req_400 – counter - info
70 Client requests received, subject to 400 errors
71 400 means we couldn't make sense of the request, it was malformed in
73 some drastic way.
74 client_req_417 – counter - info
76 Client requests received, subject to 417 errors
77 417 means that something went wrong with an Expect: header.
79 client_req – counter - info
81 Good client requests received
82 The count of parseable client requests seen.
84 cache_hit – counter - info
86 Cache hits
87 Count of cache hits. A cache hit indicates that an object has been
89 delivered to a client without fetching it from a backend server.
90 cache_hit_grace – counter - info
92 Cache grace hits
93 Count of cache hits with grace. A cache hit with grace is a cache
95 hit where the object is expired. Note that such hits are also in‐
96 cluded in the cache_hit counter.
97 cache_hitpass – counter - info
99 Cache hits for pass.
100 Count of hits for pass. A cache hit for pass indicates that Varnish
102 is going to pass the request to the backend and this decision has
103 been cached in it self. This counts how many times the cached deci‐
104 sion is being used.
105 cache_hitmiss – counter - info
107 Cache hits for miss.
108 Count of hits for miss. A cache hit for miss indicates that Varnish
110 is going to proceed as for a cache miss without request coalescing,
111 and this decision has been cached. This counts how many times the
112 cached decision is being used.
113 cache_miss – counter - info
115 Cache misses
116 Count of misses. A cache miss indicates the object was fetched from
118 the backend before delivering it to the client.
119 beresp_uncacheable – counter - info
121 Uncacheable backend responses
122 Count of backend responses considered uncacheable.
124 beresp_shortlived – counter - info
126 Shortlived objects
127 Count of objects created with ttl+grace+keep shorter than the
129 'shortlived' runtime parameter.
130 backend_conn – counter - info
132 Backend conn. success
133 How many backend connections have successfully been established.
135 backend_unhealthy – counter - info
137 Backend conn. not attempted
138 backend_busy – counter - info
140 Backend conn. too many
141 backend_fail – counter - info
143 Backend conn. failures
144 backend_reuse – counter - info
146 Backend conn. reuses
147 Count of backend connection reuses. This counter is increased when‐
149 ever we reuse a recycled connection.
150 backend_recycle – counter - info
152 Backend conn. recycles
153 Count of backend connection recycles. This counter is increased
155 whenever we have a keep-alive connection that is put back into the
156 pool of connections. It has not yet been used, but it might be, un‐
157 less the backend closes it.
158 backend_retry – counter - info
160 Backend conn. retry
161 fetch_head – counter - info
163 Fetch no body (HEAD)
164 beresp with no body because the request is HEAD.
166 fetch_length – counter - info
168 Fetch with Length
169 beresp.body with Content-Length.
171 fetch_chunked – counter - info
173 Fetch chunked
174 beresp.body with Chunked.
176 fetch_eof – counter - info
178 Fetch EOF
179 beresp.body with EOF.
181 fetch_bad – counter - info
183 Fetch bad T-E
184 beresp.body length/fetch could not be determined.
186 fetch_none – counter - info
188 Fetch no body
189 beresp.body empty
191 fetch_1xx – counter - info
193 Fetch no body (1xx)
194 beresp with no body because of 1XX response.
196 fetch_204 – counter - info
198 Fetch no body (204)
199 beresp with no body because of 204 response.
201 fetch_304 – counter - info
203 Fetch no body (304)
204 beresp with no body because of 304 response.
206 fetch_failed – counter - info
208 Fetch failed (all causes)
209 beresp fetch failed.
211 fetch_no_thread – counter - info
213 Fetch failed (no thread)
214 beresp fetch failed, no thread available.
216 pools – gauge - info
218 Number of thread pools
219 Number of thread pools. See also parameter thread_pools. NB: Pres‐
221 ently pools cannot be removed once created.
222 threads – gauge - info
224 Total number of threads
225 Number of threads in all pools. See also parameters thread_pools,
227 thread_pool_min and thread_pool_max.
228 threads_limited – counter - info
230 Threads hit max
231 Number of times more threads were needed, but limit was reached in a
233 thread pool. See also parameter thread_pool_max.
234 threads_created – counter - info
236 Threads created
237 Total number of threads created in all pools.
239 threads_destroyed – counter - info
241 Threads destroyed
242 Total number of threads destroyed in all pools.
244 threads_failed – counter - info
246 Thread creation failed
247 Number of times creating a thread failed. See VSL::Debug for diag‐
249 nostics. See also parameter thread_fail_delay.
250 thread_queue_len – gauge - info
252 Length of session queue
253 Length of session queue waiting for threads. NB: Only updates once
255 per second. See also parameter thread_queue_limit.
256 busy_sleep – counter - info
258 Number of requests sent to sleep on busy objhdr
259 Number of requests sent to sleep without a worker thread because
261 they found a busy object.
262 busy_wakeup – counter - info
264 Number of requests woken after sleep on busy objhdr
265 Number of requests taken off the busy object sleep list and resched‐
267 uled.
268 busy_killed – counter - info
270 Number of requests killed after sleep on busy objhdr
271 Number of requests killed from the busy object sleep list due to
273 lack of resources.
274 sess_queued – counter - info
276 Sessions queued for thread
277 Number of times session was queued waiting for a thread. See also
279 parameter thread_queue_limit.
280 sess_dropped – counter - info
282 Sessions dropped for thread
283 Number of times an HTTP/1 session was dropped because the queue was
285 too long already. See also parameter thread_queue_limit.
286 req_dropped – counter - info
288 Requests dropped
289 Number of times an HTTP/2 stream was refused because the queue was
291 too long already. See also parameter thread_queue_limit.
292 n_object – gauge - info
294 object structs made
295 Approximate number of HTTP objects (headers + body, if present) in
297 the cache.
298 n_vampireobject – gauge - diag
300 unresurrected objects
301 Number of unresurrected objects
303 n_objectcore – gauge - info
305 objectcore structs made
306 Approximate number of object metadata elements in the cache. Each
308 object needs an objectcore, extra objectcores are for hit-for-miss,
309 hit-for-pass and busy objects.
310 n_objecthead – gauge - info
312 objecthead structs made
313 Approximate number of different hash entries in the cache.
315 n_backend – gauge - info
317 Number of backends
318 Number of backends known to us.
320 n_expired – counter - info
322 Number of expired objects
323 Number of objects that expired from cache because of old age.
325 n_lru_nuked – counter - info
327 Number of LRU nuked objects
328 How many objects have been forcefully evicted from storage to make
330 room for a new object.
331 n_lru_moved – counter - diag
333 Number of LRU moved objects
334 Number of move operations done on the LRU list.
336 n_lru_limited – counter - info
338 Reached nuke_limit
339 Number of times more storage space were needed, but limit was
341 reached in a nuke_limit. See also parameter nuke_limit.
342 losthdr – counter - info
344 HTTP header overflows
345 s_sess – counter - info
347 Total sessions seen
348 n_pipe – gauge - info
350 Number of ongoing pipe sessions
351 pipe_limited – counter - info
353 Pipes hit pipe_sess_max
354 Number of times more pipes were needed, but the limit was reached.
356 See also parameter pipe_sess_max.
357 s_pipe – counter - info
359 Total pipe sessions seen
360 s_pass – counter - info
362 Total pass-ed requests seen
363 s_fetch – counter - info
365 Total backend fetches initiated
366 s_synth – counter - info
368 Total synthetic responses made
369 s_req_hdrbytes – counter - info
371 Request header bytes
372 Total request header bytes received
374 s_req_bodybytes – counter - info
376 Request body bytes
377 Total request body bytes received
379 s_resp_hdrbytes – counter - info
381 Response header bytes
382 Total response header bytes transmitted
384 s_resp_bodybytes – counter - info
386 Response body bytes
387 Total response body bytes transmitted
389 s_pipe_hdrbytes – counter - info
391 Pipe request header bytes
392 Total request bytes received for piped sessions
394 s_pipe_in – counter - info
396 Piped bytes from client
397 Total number of bytes forwarded from clients in pipe sessions
399 s_pipe_out – counter - info
401 Piped bytes to client
402 Total number of bytes forwarded to clients in pipe sessions
404 sess_closed – counter - info
406 Session Closed
407 sess_closed_err – counter - info
409 Session Closed with error
410 Total number of sessions closed with errors. See sc_* diag counters
412 for detailed breakdown
413 sess_readahead – counter - info
415 Session Read Ahead
416 sess_herd – counter - diag
418 Session herd
419 Number of times the timeout_linger triggered
421 sc_rem_close – counter - diag
423 Session OK REM_CLOSE
424 Number of session closes with REM_CLOSE (Client Closed)
426 sc_req_close – counter - diag
428 Session OK REQ_CLOSE
429 Number of session closes with REQ_CLOSE (Client requested close)
431 sc_req_http10 – counter - diag
433 Session Err REQ_HTTP10
434 Number of session closes with Error REQ_HTTP10 (Proto < HTTP/1.1)
436 sc_rx_bad – counter - diag
438 Session Err RX_BAD
439 Number of session closes with Error RX_BAD (Received bad req/resp)
441 sc_rx_body – counter - diag
443 Session Err RX_BODY
444 Number of session closes with Error RX_BODY (Failure receiving
446 req.body)
447 sc_rx_junk – counter - diag
449 Session Err RX_JUNK
450 Number of session closes with Error RX_JUNK (Received junk data)
452 sc_rx_overflow – counter - diag
454 Session Err RX_OVERFLOW
455 Number of session closes with Error RX_OVERFLOW (Received buffer
457 overflow)
458 sc_rx_timeout – counter - diag
460 Session Err RX_TIMEOUT
461 Number of session closes with Error RX_TIMEOUT (Receive timeout)
463 sc_rx_close_idle – counter - diag
465 Session Err RX_CLOSE_IDLE
466 Number of session closes with Error RX_CLOSE_IDLE: timeout_idle has
468 been exceeded while waiting for a client request.
469 sc_tx_pipe – counter - diag
471 Session OK TX_PIPE
472 Number of session closes with TX_PIPE (Piped transaction)
474 sc_tx_error – counter - diag
476 Session Err TX_ERROR
477 Number of session closes with Error TX_ERROR (Error transaction)
479 sc_tx_eof – counter - diag
481 Session OK TX_EOF
482 Number of session closes with TX_EOF (EOF transmission)
484 sc_resp_close – counter - diag
486 Session OK RESP_CLOSE
487 Number of session closes with RESP_CLOSE (Backend/VCL requested
489 close)
490 sc_overload – counter - diag
492 Session Err OVERLOAD
493 Number of session closes with Error OVERLOAD (Out of some resource)
495 sc_pipe_overflow – counter - diag
497 Session Err PIPE_OVERFLOW
498 Number of session closes with Error PIPE_OVERFLOW (Session pipe
500 overflow)
501 sc_range_short – counter - diag
503 Session Err RANGE_SHORT
504 Number of session closes with Error RANGE_SHORT (Insufficient data
506 for range)
507 sc_req_http20 – counter - diag
509 Session Err REQ_HTTP20
510 Number of session closes with Error REQ_HTTP20 (HTTP2 not accepted)
512 sc_vcl_failure – counter - diag
514 Session Err VCL_FAILURE
515 Number of session closes with Error VCL_FAILURE (VCL failure)
517 client_resp_500 – counter - diag
519 Delivery failed due to insufficient workspace.
520 Number of times we failed a response due to running out of workspace
522 memory during delivery.
523 ws_backend_overflow – counter - diag
525 workspace_backend overflows
526 Number of times we ran out of space in workspace_backend.
528 ws_client_overflow – counter - diag
530 workspace_client overflows
531 Number of times we ran out of space in workspace_client.
533 ws_thread_overflow – counter - diag
535 workspace_thread overflows
536 Number of times we ran out of space in workspace_thread.
538 ws_session_overflow – counter - diag
540 workspace_session overflows
541 Number of times we ran out of space in workspace_session.
543 shm_records – counter - diag
545 SHM records
546 shm_writes – counter - diag
548 SHM writes
549 shm_flushes – counter - diag
551 SHM flushes due to overflow
552 shm_cont – counter - diag
554 SHM MTX contention
555 shm_cycles – counter - diag
557 SHM cycles through buffer
558 backend_req – counter - info
560 Backend requests made
561 n_vcl – gauge - info
563 Number of loaded VCLs in total
564 n_vcl_avail – gauge - diag
566 Number of VCLs available
567 n_vcl_discard – gauge - diag
569 Number of discarded VCLs
570 vcl_fail – counter - info
572 VCL failures
573 Count of failures which prevented VCL from completing.
575 bans – gauge - info
577 Count of bans
578 Number of all bans in system, including bans superseded by newer
580 bans and bans already checked by the ban-lurker.
581 bans_completed – gauge - diag
583 Number of bans marked 'completed'
584 Number of bans which are no longer active, either because they got
586 checked by the ban-lurker or superseded by newer identical bans.
587 bans_obj – gauge - diag
589 Number of bans using obj.*
590 Number of bans which use obj.* variables. These bans can possibly
592 be washed by the ban-lurker.
593 bans_req – gauge - diag
595 Number of bans using req.*
596 Number of bans which use req.* variables. These bans can not be
598 washed by the ban-lurker.
599 bans_added – counter - diag
601 Bans added
602 Counter of bans added to ban list.
604 bans_deleted – counter - diag
606 Bans deleted
607 Counter of bans deleted from ban list.
609 bans_tested – counter - diag
611 Bans tested against objects (lookup)
612 Count of how many bans and objects have been tested against each
614 other during hash lookup.
615 bans_obj_killed – counter - diag
617 Objects killed by bans (lookup)
618 Number of objects killed by bans during object lookup.
620 bans_lurker_tested – counter - diag
622 Bans tested against objects (lurker)
623 Count of how many bans and objects have been tested against each
625 other by the ban-lurker.
626 bans_tests_tested – counter - diag
628 Ban tests tested against objects (lookup)
629 Count of how many tests and objects have been tested against each
631 other during lookup. 'ban req.url == foo && req.http.host == bar'
632 counts as one in 'bans_tested' and as two in 'bans_tests_tested'
633 bans_lurker_tests_tested – counter - diag
635 Ban tests tested against objects (lurker)
636 Count of how many tests and objects have been tested against each
638 other by the ban-lurker. 'ban req.url == foo && req.http.host ==
639 bar' counts as one in 'bans_tested' and as two in
640 'bans_tests_tested'
641 bans_lurker_obj_killed – counter - diag
643 Objects killed by bans (lurker)
644 Number of objects killed by the ban-lurker.
646 bans_lurker_obj_killed_cutoff – counter - diag
648 Objects killed by bans for cutoff (lurker)
649 Number of objects killed by the ban-lurker to keep the number of
651 bans below ban_cutoff.
652 bans_dups – counter - diag
654 Bans superseded by other bans
655 Count of bans replaced by later identical bans.
657 bans_lurker_contention – counter - diag
659 Lurker gave way for lookup
660 Number of times the ban-lurker had to wait for lookups.
662 bans_persisted_bytes – gauge - diag
664 Bytes used by the persisted ban lists
665 Number of bytes used by the persisted ban lists.
667 bans_persisted_fragmentation – gauge - diag
669 Extra bytes in persisted ban lists due to fragmentation
670 Number of extra bytes accumulated through dropped and completed bans
672 in the persistent ban lists.
673 n_purges – counter - info
675 Number of purge operations executed
676 n_obj_purged – counter - info
678 Number of purged objects
679 exp_mailed – counter - diag
681 Number of objects mailed to expiry thread
682 Number of objects mailed to expiry thread for handling.
684 exp_received – counter - diag
686 Number of objects received by expiry thread
687 Number of objects received by expiry thread for handling.
689 hcb_nolock – counter - debug
691 HCB Lookups without lock
692 hcb_lock – counter - debug
694 HCB Lookups with lock
695 hcb_insert – counter - debug
697 HCB Inserts
698 esi_errors – counter - diag
700 ESI parse errors (unlock)
701 esi_warnings – counter - diag
703 ESI parse warnings (unlock)
704 vmods – gauge - info
706 Loaded VMODs
707 n_gzip – counter - info
709 Gzip operations
710 n_gunzip – counter - info
712 Gunzip operations
713 n_test_gunzip – counter - info
715 Test gunzip operations
716 Those operations occur when Varnish receives a compressed object
718 from a backend. They are done to verify the gzip stream while it's
719 inserted in storage.
720
722 uptime – counter - info
723 Management process uptime
724 Uptime in seconds of the management process
726 child_start – counter - diag
728 Child process started
729 Number of times the child process has been started
731 child_exit – counter - diag
733 Child process normal exit
734 Number of times the child process has been cleanly stopped
736 child_stop – counter - diag
738 Child process unexpected exit
739 Number of times the child process has exited with an unexpected re‐
741 turn code
742 child_died – counter - diag
744 Child process died (signal)
745 Number of times the child process has died due to signals
747 child_dump – counter - diag
749 Child process core dumped
750 Number of times the child process has produced core dumps
752 child_panic – counter - diag
754 Child process panic
755 Number of times the management process has caught a child panic
757
759 live – gauge - debug
760 In use
761 pool – gauge - debug
763 In Pool
764 sz_wanted – gauge - debug
766 Size requested
767 sz_actual – gauge - debug
769 Size allocated
770 allocs – counter - debug
772 Allocations
773 frees – counter - debug
775 Frees
776 recycle – counter - debug
778 Recycled from pool
779 timeout – counter - debug
781 Timed out from pool
782 toosmall – counter - debug
784 Too small to recycle
785 surplus – counter - debug
787 Too many for pool
788 randry – counter - debug
790 Pool ran dry
791
793 c_req – counter - info
794 Allocator requests
795 Number of times the storage has been asked to provide a storage seg‐
797 ment.
798 c_fail – counter - info
800 Allocator failures
801 Number of times the storage has failed to provide a storage segment.
803 c_bytes – counter - info
805 Bytes allocated
806 Number of total bytes allocated by this storage.
808 c_freed – counter - info
810 Bytes freed
811 Number of total bytes returned to this storage.
813 g_alloc – gauge - info
815 Allocations outstanding
816 Number of storage allocations outstanding.
818 g_bytes – gauge - info
820 Bytes outstanding
821 Number of bytes allocated from the storage.
823 g_space – gauge - info
825 Bytes available
826 Number of bytes left in the storage.
828
830 c_req – counter - info
831 Allocator requests
832 Number of times the storage has been asked to provide a storage seg‐
834 ment.
835 c_fail – counter - info
837 Allocator failures
838 Number of times the storage has failed to provide a storage segment.
840 c_bytes – counter - info
842 Bytes allocated
843 Number of total bytes allocated by this storage.
845 c_freed – counter - info
847 Bytes freed
848 Number of total bytes returned to this storage.
850 g_alloc – gauge - info
852 Allocations outstanding
853 Number of storage allocations outstanding.
855 g_bytes – gauge - info
857 Bytes outstanding
858 Number of bytes allocated from the storage.
860 g_space – gauge - info
862 Bytes available
863 Number of bytes left in the storage.
865
867 c_req – counter - info
868 Allocator requests
869 Number of times the storage has been asked to provide a storage seg‐
871 ment.
872 c_fail – counter - info
874 Allocator failures
875 Number of times the storage has failed to provide a storage segment.
877 c_bytes – counter - info
879 Bytes allocated
880 Number of total bytes allocated by this storage.
882 c_freed – counter - info
884 Bytes freed
885 Number of total bytes returned to this storage.
887 g_alloc – gauge - info
889 Allocations outstanding
890 Number of storage allocations outstanding.
892 g_bytes – gauge - info
894 Bytes outstanding
895 Number of bytes allocated from the storage.
897 g_space – gauge - info
899 Bytes available
900 Number of bytes left in the storage.
902 g_smf – gauge - info
904 N struct smf
905 g_smf_frag – gauge - info
907 N small free smf
908 g_smf_large – gauge - info
910 N large free smf
911
913 happy – bitmap - info
914 Happy health probes
915 bereq_hdrbytes – counter - info
917 Request header bytes
918 Total backend request header bytes sent
920 bereq_bodybytes – counter - info
922 Request body bytes
923 Total backend request body bytes sent
925 beresp_hdrbytes – counter - info
927 Response header bytes
928 Total backend response header bytes received
930 beresp_bodybytes – counter - info
932 Response body bytes
933 Total backend response body bytes received
935 pipe_hdrbytes – counter - info
937 Pipe request header bytes
938 Total request bytes sent for piped sessions
940 pipe_out – counter - info
942 Piped bytes to backend
943 Total number of bytes forwarded to backend in pipe sessions
945 pipe_in – counter - info
947 Piped bytes from backend
948 Total number of bytes forwarded from backend in pipe sessions
950 conn – gauge - info
952 Concurrent connections used
953 The number of currently used connections to the backend. This number
955 is always less or equal to the number of connections to the backend
956 (as, for example shown as ESTABLISHED for TCP connections in net‐
957 stat) due to connection pooling.
958 req – counter - info
960 Backend requests sent
961 unhealthy – counter - info
963 Fetches not attempted due to backend being unhealthy
964 busy – counter - info
966 Fetches not attempted due to backend being busy
967 Number of times the max_connections limit was reached
969 fail – counter - info
971 Connections failed
972 Counter of failed opens. Detailed reasons are given in the fail_*
974 counters (DIAG level) and in the log under the FetchError tag.
975 This counter is the sum of all detailed fail_* counters.
977 All fail_* counters may be slightly inaccurate for efficiency.
979 fail_eacces – counter - diag
981 Connections failed with EACCES or EPERM
982 fail_eaddrnotavail – counter - diag
984 Connections failed with EADDRNOTAVAIL
985 fail_econnrefused – counter - diag
987 Connections failed with ECONNREFUSED
988 fail_enetunreach – counter - diag
990 Connections failed with ENETUNREACH
991 fail_etimedout – counter - diag
993 Connections failed ETIMEDOUT
994 fail_other – counter - diag
996 Connections failed for other reason
997 helddown – counter - diag
999 Connection opens not attempted
1000 Connections not attempted during the backend_local_error_holddown or
1002 backend_remote_error_holddown interval after a fundamental connec‐
1003 tion issue.
1004
1006 Counters which track the activity in the different classes of mu‐
1007 tex-locks.
1008 The counts may be slightly wrong if there are more than one lock in‐
1010 stantiated in each class (ie: .creat > 1)
1011 creat – counter - debug
1013 Created locks
1014 destroy – counter - debug
1016 Destroyed locks
1017 locks – counter - debug
1019 Lock Operations
1020 dbg_busy – counter - debug
1022 Contended lock operations
1023 If the lck debug bit is set: Lock operations which returned EBUSY on
1025 the first locking attempt.
1026 If the lck debug bit is unset, this counter will never be incre‐
1028 mented even if lock operations are contended.
1029 dbg_try_fail – counter - debug
1031 Contended trylock operations
1032 If the lck debug bit is set: Trylock operations which returned
1034 EBUSY.
1035 If the lck debug bit is unset, this counter will never be incre‐
1037 mented even if lock operations are contended.
1038
1040 This man page was written by Lasse Karstensen, using content from
1041 vsc2rst written by Tollef Fog Heen.
1042 VARNISH-COUNTERS(7)