> On 29/12/2021 11:20 tobisworld at gmail.com wrote: > > > Hi list > > I have weird issue with my Dovecot 2.3.17.1 (476cd46418) > When deleting a certain amount of messages from my INBOX via my MUA > (Evolution) all over sudden dovecot starts to panic > > Panic: file mail-index-transaction-finish.c: line 185 > (mail_index_transaction_get_uid): assertion failed: (seq <= t->view- > >map->hdr.messages_count) > > imap(REDACTED)<24075><P0/vLkXUNt4KQmMB>: Error: Raw backtrace: > /usr/lib64/dovecot/libdovecot.so.0(backtrace_append+0x42) > [0x7f09274d4142] -> > /usr/lib64/dovecot/libdovecot.so.0(backtrace_get+0x1e) [0x7f09274d424e] > -> /usr/lib64/dovecot/libdovecot.so.0(+0xf72fe) [0x7f09274e22fe] -> > /usr/lib64/dovecot/libdovecot.so.0(+0xf73a1) [0x7f09274e23a1] -> > /usr/lib64/dovecot/libdovecot.so.0(i_fatal+0) [0x7f0927430e38] -> > /usr/lib64/dovecot/libdovecot-storage.so.0(+0xf438d) [0x7f092789d38d] - > > /usr/lib64/dovecot/libdovecot-storage.so.0(+0xf43da) [0x7f092789d3da] > -> /usr/lib64/dovecot/libdovecot- > storage.so.0(mail_index_transaction_finish+0x177) [0x7f092789da87] -> > /usr/lib64/dovecot/libdovecot-storage.so.0(+0xf9ad6) [0x7f09278a2ad6] - > > /usr/lib64/dovecot/libdovecot- > storage.so.0(mail_index_transaction_commit_full+0x9d) [0x7f09278a2fad] > -> /usr/lib64/dovecot/libdovecot- > storage.so.0(mail_index_transaction_commit+0x1c) [0x7f09278a306c] -> > /usr/lib64/dovecot/libdovecot- > storage.so.0(mail_cache_set_seq_corrupted_reason+0x61) [0x7f092788c921] > -> /usr/lib64/dovecot/libdovecot- > storage.so.0(mail_set_mail_cache_corrupted+0xed) [0x7f09277fbbbd] -> > /usr/lib64/dovecot/libdovecot-storage.so.0(+0x7a3e9) [0x7f09278233e9] - > > /usr/lib64/dovecot/libdovecot-storage.so.0(+0x7a6b0) [0x7f09278236b0] > -> /usr/lib64/dovecot/libdovecot- > storage.so.0(mail_get_physical_size+0x38) [0x7f09277e8ba8] -> > /usr/libexec/dovecot//lib20_mail_log_plugin.so(+0x2465) > [0x7f09263dd465] -> > /usr/libexec/dovecot//lib15_notify_plugin.so(notify_contexts_mail_copy+ > 0x5c) [0x7f09265e13ec] -> > /usr/libexec/dovecot//lib15_notify_plugin.so(+0x25d4) [0x7f09265e25d4] > -> /usr/libexec/dovecot//lib10_quota_plugin.so(+0x102ae) > [0x7f09269f92ae] -> /usr/libexec/dovecot//lib01_acl_plugin.so(+0xef3e) > [0x7f0926c10f3e] -> /usr/lib64/dovecot/libdovecot- > storage.so.0(+0x50c4c) [0x7f09277f9c4c] -> > /usr/lib64/dovecot/libdovecot-storage.so.0(mailbox_move+0x58) > [0x7f09277f9ec8] -> dovecot/imap(+0x1175b) [0x55a2c0ea575b] -> > dovecot/imap(command_exec+0x64) [0x55a2c0eb3f04] -> > dovecot/imap(+0x1dde2) [0x55a2c0eb1de2] -> dovecot/imap(+0x1de71) > [0x55a2c0eb1e71] -> dovecot/imap(+0x1e039) [0x55a2c0eb2039] -> > dovecot/imap(client_handle_input+0x205) [0x55a2c0eb22f5] -> > dovecot/imap(client_input+0x79) [0x55a2c0eb28d9] > > imap(REDACTED)<24075><P0/vLkXUNt4KQmMB>: Fatal: master: service(imap): > child 24075 killed with signal 6 (core dumps disabled - > https://dovecot.org/bugreport.html#coredumps) > > > From that point on ANY access to INBOX from my MUA causes the error. > The only way that I found to get around is to unsubscribe the INBOX > folder and subscribe again. Then the error is gone until I delete a > bunch of messages again. > The error only occurs if I delete the messages within short time or > select multiple (about 20msg at once). If I delete them one-by-one > always waiting for expunge message in dovecot logs to appear then I can > delete as many as I want. > > I have not verified with other MUA but in Evolution I can reproduce the > issue. > > Cheers and have a good one > > tobiDo you have multiple backends with concurrent access via e.g. nfs? Aki
tobisworld at gmail.com
2021-Dec-29 09:40 UTC
Panic: file mail-index-transaction-finish.c: line 185
On Wed, 2021-12-29 at 11:26 +0200, Aki Tuomi wrote:> > > Do you have multiple backends with concurrent access via e.g. nfs? > > Aki >Aki, yes all mailboxes are in a nfs share. Concurrent access could be possible as I have several MUA (my desktop and my mobiles). Although only the client on my desktop (evolution) is currently doing something. My mobiles are "sleeping" atm. Do you think the nfs share could be the culprit? Cheers tobi
Arkadiusz Miśkiewicz
2022-Sep-15 04:57 UTC
Panic: file mail-index-transaction-finish.c: line 185
On 29.12.2021 10:26, Aki Tuomi wrote:> >> On 29/12/2021 11:20 tobisworld at gmail.com wrote: >> >> >> Hi list >> >> I have weird issue with my Dovecot 2.3.17.1 (476cd46418) >> When deleting a certain amount of messages from my INBOX via my MUA >> (Evolution) all over sudden dovecot starts to panic >> >> Panic: file mail-index-transaction-finish.c: line 185 >> (mail_index_transaction_get_uid): assertion failed: (seq <= t->view- >>> map->hdr.messages_count) >> >> imap(REDACTED)<24075><P0/vLkXUNt4KQmMB>: Error: Raw backtrace: >> /usr/lib64/dovecot/libdovecot.so.0(backtrace_append+0x42) >> [0x7f09274d4142] -> >> /usr/lib64/dovecot/libdovecot.so.0(backtrace_get+0x1e) [0x7f09274d424e] >> -> /usr/lib64/dovecot/libdovecot.so.0(+0xf72fe) [0x7f09274e22fe] -> >> /usr/lib64/dovecot/libdovecot.so.0(+0xf73a1) [0x7f09274e23a1] -> >> /usr/lib64/dovecot/libdovecot.so.0(i_fatal+0) [0x7f0927430e38] ->I also sometimes see this on 2.3.19.1: Sep 15 05:05:43 mbox dovecot: imap(marek): pid=<14897> session=<JOBjha7oCsVR1Y7+>, Panic: file mail-index-transaction-finish.c: line 185 (mail_index_transaction_get_uid): assertion failed: (seq <= t->view->map->hdr.messages_count) Sep 15 05:05:43 mbox dovecot: imap(marek): pid=<14897> session=<JOBjha7oCsVR1Y7+>, Error: Raw backtrace: #0 t_askpass[0x7f16bf8658e0] -> #1 backtrace_append[0x7f16bf865b50] -> #2 backtrace_get[0x7f16bf865cb0] -> #3 i_syslog_error_handler[0x7f16bf8727d0] -> #4 i_syslog_fatal_handler[0x7f16bf872900] -> #5 i_panic[0x7f16bf7c62d6] -> #6 mail_index_sync_set_corrupted[0x7f16bf996d27] -> #7 mail_transaction_expunge_guid_cmp[0x7f16bfa43fe0] -> #8 mail_index_transaction_finish[0x7f16bfa44550] -> #9 mail_index_transaction_unref[0x7f16bfa48c30] -> #10 mail_index_transaction_commit_full[0x7f16bfa49110] -> #11 mail_index_transaction_commit[0x7f16bfa491f0] -> #12 mail_cache_set_seq_corrupted_reason[0x7f16bf993a4f] -> #13 mail_set_mail_cache_corrupted[0x7f16bf9ae690] -> #14 maildir_keywords_idx_char[0x7f16bf9d2a50] -> #15 maildir_keywords_idx_char[0x7f16bf9d2de0] -> #16 mail_get_physical_size[0x7f16bf99b770] -> #17 [unw_get_proc_name() failed: -10] -> #18 notify_contexts_mail_copy[0x7f16bead94b0] -> #19 notify_plugin_deinit[0x7f16beada440] -> #20 quota_plugin_deinit[0x7f16bf4b9350] -> #21 acl_mailbox_right_lookup[0x7f16bf4d7720] -> #22 mailbox_save_begin[0x7f16bf9ac880] -> #23 mailbox_copy[0x7f16bf9aca00] -> #24 cmd_close[0x55978a0b0980] -> #25 command_exec[0x55978a0bf220] -> #26 client_handle_unfinished_cmd[0x55978a0bd2b0] -> #27 client_handle_unfinished_cmd[0x55978a0bd2b0] -> #28 client_handle_unfinished_cmd[0x55978a0bd2b0] -> #29 client_handle_input[0x55978a0bd630] -> #30 client_input[0x55978a0bdca0] -> #31 io_loop_call_io[0x7f16bf888850] -> #32 io_loop_handler_run_internal[0x7f16bf889e90] -> #33 io_loop_handler_run[0x7f16bf888910] -> #34 io_loop_run[0x7f16bf888ae0] -> #35 master_service_run[0x7f16bf7fbe70] -> #36 main[0x55978a0ae9f0] -> #37 __libc_init_first[0x7f16bf5a34d0] -> #38 __libc_start_main[0x7f16bf5a3580] -> #39 _start[0x55978a0aefa0] Sep 15 05:05:43 mbox dovecot: imap(marek): pid=<14897> session=<JOBjha7oCsVR1Y7+>, Fatal: master: service(imap): child 14897 killed with signal 6 (core dumps disabled - https://dovecot.org/bugreport.html#coredumps) No NFS involved here (linux + xfs). -- Arkadiusz Mi?kiewicz, arekm / ( maven.pl | pld-linux.org )