• Wrong benchmark code, do not run several benchmarks at once!

    I installed benchmark on one of our websites which runs on the same physical server (And IP) as our production but with a lot less plugins and tools etc. I ran it and can continue to run it fine on this small website.

    I want to migrate it to main website but I keep getting an error as follows. I was able to run it once before installing on the small website (which I did to see if there was much of a difference in results)
    Wrong benchmark code, do not run several benchmarks at once!

    I have tried to deactivate the plugin, uninstall it and reinstall. same thing.

    Secondly, I get result on both websites as “Importing large amount of data to database 0.98”, my big site uses ludicrousDB and reads from 2 and writes to 1 DB server. The servers are physically connected with 10Gb so wondered why this result. I get an overall result of 8, its just the DB I wondered if this was real bad, or was possible due to the DB server off the host adding a 0.2ms latency.

    Wonder if you can make any recommendations, running on ZFS with small byte size block

    # Memory and buffer settings
    innodb_buffer_pool_size = 24G
    innodb_log_file_size = 1G
    innodb_log_buffer_size = 64M
    innodb_flush_method = O_DIRECT
    innodb_flush_log_at_trx_commit = 1
    innodb_file_per_table = 1

    # Thread settings
    innodb_thread_concurrency = 16
    innodb_read_io_threads = 16
    innodb_write_io_threads = 16
    innodb_io_capacity = 2000
    innodb_io_capacity_max = 4000

    # AIO settings
    innodb_use_native_aio = 1

    # Other settings
    slow_query_log = 1
    slow_query_log_file = /var/log/mysql/mysql-slow.log
    long_query_time = 2

    # Timezone
    default_time_zone = 'Pacific/Auckland'

    max_binlog_size = 100M
    expire_logs_days = 7
    relay_log_space_limit = 1G


    # Allow infile for IPMS
    local-infile=1


    [mysqld]
    disabled_storage_engines="MyISAM,BLACKHOLE,FEDERATED,ARCHIVE,MEMORY"
    gtid_mode = ON
    enforce_gtid_consistency = ON
    master_info_repository = TABLE
    relay_log_info_repository = TABLE
    binlog_checksum = NONE
    log_replica_updates = ON
    log_bin = binlog
    binlog_format = ROW
    sync_binlog = 1
    binlog_expire_logs_seconds = 3600 # Purge binlogs after 1 hour (adjust as needed)
    binlog_row_image = MINIMAL
    transaction_write_set_extraction = XXHASH64
Viewing 1 replies (of 1 total)
  • Plugin Author Anton Aleksandrov

    (@antonaleksandrov)

    Hey,
    Thanks for pointing this out.. I have added a call to clear “running benchmark” flag. Just update and save your settings and you should be good..

    Regarding database speed. Well, there’s not single answer. Complex cluster or replicated solutions can be beneficial, but for “quick insert/quite read” might give you a bit unexpected results.. Same if you have some DB throttling. Main question would be – is it fast or slow? I am not aware of DB engine you use.. :/

Viewing 1 replies (of 1 total)

The topic ‘Wrong benchmark code, do not run several benchmarks at once!’ is closed to new replies.