Database cache corrupted

[sebelk]

Hi,

Sometimes zabbix goes zombie, for example deleting items from templates, or linking/unlinking templates with hosts....

zabbix log looks like this:

3579:20080424:102742 Unsupported history value type 1083500544. Database cache corrupted?
3543:20080424:102742 One child process died. Exiting

---

What does mean "database cache corrupted"? it can be fixed? (I am using MySQL with InnoDB)



Thanks in advance..

[Kees Jan Koster]

I had some innodb tables go corrupt on me. I replaced them with myisam tables and everything is peachy since.

[sebelk]

Thanks but I'd want to keep using InnoDB

[sebelk]

Same problem zabbix goes zombie :'-(

Still I've got stuck with this problem, I use a script that restart zabbix when it goes zombie, but it's only a poor and temporary workaround

5442:20080425:142639 Unsupported history value type 1072693248. Database cache corrupted?
5411:20080425:142639 One child process died. Exiting ...
5411:20080425:142641 In free_database_cache()
5411:20080425:142641 In DCsync_all(items 10573 pool:trends 43 pool:history:39921)

I don't understand what's wrong, this is my config files:

/etc/zabbix/zabbix_server.conf:

StartPollers=5
SenderFrequency=300
DebugLevel=3
Timeout=5
PidFile=/var/run/zabbix/zabbix.pid
LogFile=/var/log/zabbix/zabbix_server-1.5.log
AlertScriptsPath=/var/lib/zabbix/
ExternalScripts=/etc/zabbix-1.5/externalscripts
DBName=zabbix15
DBUser=zabbix
DBPassword=uurururu

eof

and my.cnf:

[client]
port = 3306
socket = /var/lib/mysql/mysql.sock
[mysqld]
port = 3306
socket = /var/lib/mysql/mysql.sock
back_log = 50
max_connections = 100
max_connect_errors = 10
table_cache = 1024
max_allowed_packet = 16M
binlog_cache_size = 1M
max_heap_table_size = 128M
sort_buffer_size = 8M
join_buffer_size = 3M
thread_cache_size = 8
thread_concurrency = 8
query_cache_size = 256M
query_cache_limit = 16M
ft_min_word_len = 4
default_table_type = InnoDB
thread_stack = 192K
transaction_isolation = REPEATABLE-READ
tmp_table_size = 64M
log_slow_queries = /var/log/mysqld/slow-query-log
long_query_time = 5
log_long_format
tmpdir = /tmp
log_queries_not_using_indexes = /var/log/mysqld/not-indexes.log
expire_logs_days = 2
server-id = 1
key_buffer_size = 8M
read_buffer_size = 2M
read_rnd_buffer_size = 16M
bulk_insert_buffer_size = 64M
myisam_sort_buffer_size = 128M
myisam_max_sort_file_size = 10G
myisam_max_extra_sort_file_size = 10G
myisam_repair_threads = 1
myisam_recover
skip-bdb
innodb_additional_mem_pool_size = 16M
innodb_buffer_pool_size = 600M
innodb_data_file_path = ibdata1:128M;ibdata2:50M:autoextend:max:12800M
innodb_file_io_threads = 4
innodb_thread_concurrency = 16
innodb_flush_log_at_trx_commit = 1
innodb_log_buffer_size = 8M
innodb_log_file_size = 256M
innodb_log_files_in_group = 3
innodb_max_dirty_pages_pct = 90
innodb_lock_wait_timeout = 120
[mysqldump]
quick
max_allowed_packet = 16M
[mysql]
no-auto-rehash
[isamchk]
key_buffer = 512M
sort_buffer_size = 512M
read_buffer = 8M
write_buffer = 8M
[myisamchk]
key_buffer = 512M
sort_buffer_size = 512M
read_buffer = 8M
write_buffer = 8M
[mysqlhotcopy]
interactive-timeout
[mysqld_safe]
open-files-limit = 8192

eof

thanks in advance

[Kees Jan Koster]

In my case the MySQL server reported the table corruption and I tried to use the REPAIR TABLE command, but that failed too.

My solution was to mysqldump the broken tables and drop them from the database. I then changed the CREATE TABLE statement in the dump from MyISAM to InnoDB and I have not seen corruption since.

Kees Jan

[sebelk]

In my case the MySQL server reported the table corruption and I tried to use the REPAIR TABLE command, but that failed too.

Thanks, but my problem is with database cache....

My solution was to mysqldump the broken tables and drop them from the database. I then changed the CREATE TABLE statement in the dump from MyISAM to InnoDB and I have not seen corruption since.

Kees Jan

Tables from database already are using InnoDB engine

Plz, please, plzzzz! what can I do in order to prevent zombie state and database cache corruption ?