lots of APPEND error finding folder

romeor

Hello,

I’m seeing pretty much APPEND error messages during syncing accounts using both kopano-migration-imap and latest built imapsync (one is a fork of anohter?).

My system is:
KVM virtualized Debian GNU/Linux 10
Kopano Core:
11.0.2
WebApp:
6.0.0.66-1+2030.1

VM specs:
RAM 16GB
cpu-cores 6 Intel® Xeon® Gold 6326 CPU @ 2.90GHz (for guest)
disk: virtio stored at qemu img file.
network: virtio

Disk system on host: SSD raid5

Messages look like:

- msg Sent/160890 {939} could not append ( Subject:[Re: ****], Date:["03-Dec-2012 14:09:47 +0000"], Size:[939], Flags:[\Seen] ) to folder Sent: 4157 NO [TRYCREATE] APPEND error finding folder
- msg Sent/160891 {1411} could not append ( Subject:[****], Date:["04-Dec-2012 08:50:23 +0000"], Size:[1411], Flags:[\Seen] ) to folder Sent: 4158 NO [TRYCREATE] APPEND error finding folder
- msg Sent/160892 {1059} could not append ( Subject:[[Fwd: K****]], Date:["04-Dec-2012 11:05:56 +0000"], Size:[1059], Flags:[\Seen] ) to folder Sent: 4159 NO [TRYCREATE] APPEND error finding folder

server.conf

enable_sql_procedures = yes
threads = 12
cache_cell_size = 3750
cache_object_size = 16M
cache_quota_size = 1M
cache_quota_lifetime = 1
cache_acl_size = 1M
cache_store_size = 1M
cache_userdetails_size = 25M
cache_indexedobject_size = 32M

[mysqld]

ignore_db_dirs = lost+found
thread_stack = 512K
transaction-isolation = READ-COMMITTED
max_connections = 100
max_allowed_packet = 32M
tmp_table_size = 64M
max_heap_table_size = 64M
sort_buffer_size = 2M
join_buffer_size = 2M
thread_cache_size = 21
innodb_write_io_threads = 10
innodb_read_io_threads = 10
table_open_cache = 500
query_cache_size = 64M
query_cache_limit = 2M
innodb_buffer_pool_size = 3G
innodb_autoextend_increment = 1000
innodb_file_per_table = OFF
innodb_flush_log_at_trx_commit = 2
innodb_log_buffer_size = 8M
innodb_log_file_size = 100M
innodb_log_files_in_group = 3
innodb_lock_wait_timeout = 120
innodb_flush_method = O_DSYNC

gateway.conf

pop3_listen = :110
pop3s_listen = :995
imap_listen = :143
imaps_listen = :993
ssl_private_key_file = /etc/kopano/imap.key
ssl_certificate_file = /etc/kopano/imap.pem


log_method = file
log_file = /var/log/kopano/gateway.log

Am personally suspecting innodb_flush_method in my.conf. Could it be a reason, if am runnign mysql on the same virtualized host? Maybe it is too fast for it? APPEND are errors meaning host2, so it seems like it writes too fast or can’t flush as fast as it would like?

Some more information: if i start syncing again - it syncs mails with erros, but then errors come with other mails. So if I repeat syncing a lot of times, I get boxes synced. I followed the https://imapsync.lamiral.info/FAQ.d/FAQ.APPEND_errors.txt and nothing helped. So my guess is about errors on writes to the disk or in database?

romeor

never mind. it seems that my mysql database was corrupted after I changed innodb logfiles path… it statrted to become worse and worse and today I was not even able to read from some boxes. So I ended up with new installation, seems working fine now, but I decided to go with kopano-core package provided by debian 10 buster for easier update procedures in the future.