Help with Percona configuration

dbembibre

Active member
I have a dedicated DB server, that is not swapping but i think that can be faster. Is a dedicated DB server with Percona 5.5 and Elastic search installed on it. The JVM is configured to take 512MB

8 Cores
Code:
Intel(R) Xeon(R) CPU          E5410  @ 2.33GHz

16 GB of Ram

SCSI disk 15RPM
Code:
[root@back daniel]# /usr/sbin/lshw -class disk -class storage
  *-storage
      description: RAID bus controller
      product: MegaRAID SAS 1078
      vendor: LSI Logic / Symbios Logic
      physical id: 0
      bus info: pci@0000:01:00.0
      logical name: scsi0
      version: 04
      width: 64 bits
      clock: 33MHz
      capabilities: storage pciexpress msi msix pm vpd bus_master cap_list
      configuration: driver=megaraid_sas latency=0 module=megaraid_sas
    *-disk
          description: SCSI Disk
          product: PERC 6/i
          vendor: DELL
          physical id: 2.0.0
          bus info: scsi@0:2.0.0
          logical name: /dev/sda
          version: 1.11
          serial: 002822d8031d72191000e5ead490ec01
          size: 136GiB (146GB)
          capabilities: partitioned partitioned:dos
          configuration: ansiversion=5 signature=00000080

My mysql configuration

Code:
max_allowed_packet            = 16M
max_connect_errors            = 1000000
skip_name_resolve
max_connections = 500

#MYISAM
key_buffer_size                = 32M
myisam_recover                = FORCE,BACKUP

# SAFETY #
max_allowed_packet            = 16M
max_connect_errors            = 1000000
skip_name_resolve
back_log = 50
max_connections = 300

# CACHES AND LIMITS #
join_buffer_size              = 64M
read_buffer_size              = 2M
sort_buffer_size              = 8M
tmp_table_size                = 256M
max_heap_table_size            = 256M
query_cache_type              = 1
query_cache_size              = 64M
thread_cache_size              = 512
open_files_limit              = 65535
table_definition_cache        = 8000
table_open_cache              = 8000


expire_logs_days              = 14
sync_binlog                    = 1

#INNODB
innodb_read_ahead = none

innodb_thread_concurrency = 4
innodb_flush_method  = O_DIRECT

innodb_log_files_in_group  = 2
innodb_log_file_size  = 256M
innodb_flush_log_at_trx_commit = 2
innodb_file_per_table  = 1

innodb_lock_wait_timeout=50
innodb_log_buffer_size=8M
innodb_buffer_pool_size  = 10G
innodb_additional_mem_pool_size = 32M

innodb_io_capacity = 2000
innodb_read_io_threads = 4
innodb_write_io_threads = 2
 
Percona config looks OK TBH.

Would probably see a decent benefit swapping the 15K disk to an SSD though.
 
Percona config looks OK TBH.

Would probably see a decent benefit swapping the 15K disk to an SSD though.

Yes an SSD can help a lot, but i think that i found my problem. I have max_connections to 600 i have around 1000 users online, but mysql show that max connections was reached. Max connections opened 601.

Aborted clients 599
Aborted connects 1.2 k

I only have installed xentrader, enhanced search, xenresources, tapatalk and sitemap, any idea ?
I think that one of this addons is not closed properly the mysql connections
 
Back
Top Bottom