dovecot - Jul 2019 - dovecot.index.log: duplicate transaction log sequence (3)

Hello,

I am running Dovecot 2.3.5.1 on OpenBSD 6.5 with RainLoop as IMAP webmail client
and just noticed the following error messages about duplicate transaction log
sequences in the index log:

Jul 01 13:15:58 Error:
imap(<REMOVED_EMAIL1>)<21324><TVSmwJyMd0C5D+Vb>: Transaction
log /var/vmail/<REMOVED_DOMAIN>/<REMOVED_EMAIL1>/dovecot.index.log:
duplicate transaction log sequence (3)
Jul 02 16:33:35 Error:
imap(<REMOVED_EMAIL2>)<6812><gn4yobOMJlW5D+Vb>: Transaction
log
/var/vmail/<REMOVED_DOMAIN>/<REMOVED_EMAIL2>/dovecot.list.index.log:
duplicate transaction log sequence (2)
Jul 02 16:33:35 Panic:
imap(<REMOVED_EMAIL2>)<6812><gn4yobOMJlW5D+Vb>: file
mail-index-write.c: line 137 (mail_index_write): assertion failed:
(file->hdr.prev_file_seq == hdr->log_file_seq)
Jul 02 16:33:35 Fatal:
imap(<REMOVED_EMAIL2>)<6812><gn4yobOMJlW5D+Vb>: master:
service(imap): child 6812 killed with signal 6 (core not dumped -
https://dovecot.org/bugreport.html#coredumps - set service imap {
drop_priv_before_exec=yes })

Could this problem occur because I am using a Linux NFS server to store all the
mails mounted under /var/mail ?

Any hints what I can do to avoid these error messages?

Thank you in advance.

Best regards,
Mabi

> 
>  
> Hello,
> 
> I am running Dovecot 2.3.5.1 on OpenBSD 6.5 with RainLoop as IMAP webmail
client and just noticed the following error messages about duplicate transaction
log sequences in the index log:
> 
> Jul 01 13:15:58 Error:
imap(<REMOVED_EMAIL1>)<21324><TVSmwJyMd0C5D+Vb>: Transaction
log /var/vmail/<REMOVED_DOMAIN>/<REMOVED_EMAIL1>/dovecot.index.log:
duplicate transaction log sequence (3)
> Jul 02 16:33:35 Error:
imap(<REMOVED_EMAIL2>)<6812><gn4yobOMJlW5D+Vb>: Transaction
log
/var/vmail/<REMOVED_DOMAIN>/<REMOVED_EMAIL2>/dovecot.list.index.log:
duplicate transaction log sequence (2)
> Jul 02 16:33:35 Panic:
imap(<REMOVED_EMAIL2>)<6812><gn4yobOMJlW5D+Vb>: file
mail-index-write.c: line 137 (mail_index_write): assertion failed:
(file->hdr.prev_file_seq == hdr->log_file_seq)
> Jul 02 16:33:35 Fatal:
imap(<REMOVED_EMAIL2>)<6812><gn4yobOMJlW5D+Vb>: master:
service(imap): child 6812 killed with signal 6 (core not dumped -
https://dovecot.org/bugreport.html#coredumps - set service imap {
drop_priv_before_exec=yes })
> 
> Could this problem occur because I am using a Linux NFS server to store all
the mails mounted under /var/mail ?
> 
> Any hints what I can do to avoid these error messages?
> 
> Thank you in advance.
> 
> Best regards,
> Mabi
Are you by chance accessing mails using two different dovecot instances? This is
not supported to be done concurrently, you need to use some sort of solution,
such as dovecot director, to ensure user lands to one backend at a time.

Aki

??????? Original Message ???????
On Tuesday, July 2, 2019 6:39 PM, Aki Tuomi <aki.tuomi at
open-xchange.com> wrote:
> Are you by chance accessing mails using two different dovecot instances?
This is not supported to be done concurrently, you need to use some sort of
solution, such as dovecot director, to ensure user lands to one backend at a
time.
If you mean two different servers by asking if I access mails "using two
different dovecot instances" then the answer is no: I have everything one
one server. Does this answer your question properly?

On the config side I have also made sure I have the "mmap_disable =
yes" parameter for NFS. Do I maybe need to tune some additional parameters?