mypsl 0.32

安装

通过 PIP： pip install mypsl

用法

usage: mypsl [-h] [-H HOST] [-p PORT] [-u USER] [-P PASSWD] [-ch CHARSET]
     [--config CONNECT_CONFIG] [-sm SALT_MINION]
     [-l LOOP_SECOND_INTERVAL] [-lm LOOP_MAX] [-dft] [-c COMMAND]
     [-s STATE] [-t TIME] [-d DATABASE] [-q QUERY] [-i] [-isr]
     [--debug] [-o ORDER_BY] [-T] [-v] [--slack] [--kill]
     [-kt KILL_THRESHOLD] [-ka] [-ky] [-kl KILL_LOG] [-klo]

(mypsl: mypsl: 0.19) :: MySQL Process list watcher & query
killer.

optional arguments:
  -h, --help            show this help message and exit

Connection Options:
  -H HOST, --host HOST  The host to get the process list from. If localhost,
                        we will attempt to find and use the socket file first.
                        (default: localhost)
  -p PORT, --port PORT  The host's port. If the host is localhost, we will
                        attempt to find and use the socket file first.
                        (default: 3306)
  -u USER, --user USER  The user to connect to the host as. (default: root)
  -P PASSWD, --pass PASSWD
                        The password for authentication. (default: )
  -ch CHARSET, --charset CHARSET
                        Charset to use with the database. (default: utf8)
  --config CONNECT_CONFIG
                        Load connection configuration from a file in
                        /root/.mypsl. Just provide the filename. This will
                        override any other connection information provided
                        (default: None)
  -sm SALT_MINION, --salt-minion SALT_MINION
                        Connect to mysql running on a salt minion. Do not use
                        any other connection options with this.
                        mysql:connection:user and mysql:connection:pass must
                        exist in pillar data. (default: None)

Configuration Options:
  -l LOOP_SECOND_INTERVAL, --loop LOOP_SECOND_INTERVAL
                        Time in seconds between getting the process list.
                        (default: 0)
  -lm LOOP_MAX, --loop_max LOOP_MAX
                        If specified we will stop execution when the loop
                        interval reaches this number (default: None)
  -dft, --default       Run with defaults. Loop interval: 3 seconds, command
                        like query or connect, order by time asc, id asc,
                        truncate query to 1000. (default: False)
  -c COMMAND, --command COMMAND
                        Lookup processes running as this command. (default:
                        None)
  -s STATE, --state STATE
                        Lookup processes running in this state. (default:
                        None)
  -t TIME, --time TIME  Lookup processes running longer than the specified
                        time in seconds. (default: None)
  -d DATABASE, --database DATABASE
                        Lookup processes running against this database.
                        (default: None)
  -q QUERY, --query QUERY
                        Lookup processes where the query starts with this
                        specification. (default: None)
  -i, --id              Only print back the ID of the processes. (default:
                        False)
  -isr, --ignore_system_user
                        Ignore the 'system user' (default: False)
  --debug               Provide debug output. (default: False)
  -o ORDER_BY, --order_by ORDER_BY
                        Order the results by a particular column: "user", "db
                        asc", "db desc", "time desc"...etc (default: None)
  -T, --trim_info       Trim the info field (the query) to 1000 (default:
                        False)
  -v, --version         Show the installed program version and quit. (default:
                        False)
  --slack               Send a notification to slack anytime we kill a query.
                        Slack Auth/Config file must exist @ ~/.slack_auth
                        (default: False)

Kill Options:
  --kill                Kill the queries that we find. (default: False)
  -kt KILL_THRESHOLD, --kill_threshold KILL_THRESHOLD
                        The kill threshold. If a number is provided, we'll
                        need to hit that many total connections before killing
                        queries. You can set this to 'off' as well, which
                        kills queries no matter how many connections there
                        are. (default: 100)
  -ka, --kill_all       If this flag is provided, we'll attempt to kill
                        everything, not only select queries. Use with
                        caution! (default: False)
  -ky, --kill_yes       If this is provided we won't stop to ask if you are
                        sure that you want to kill queries. (default: False)
  -kl KILL_LOG, --kill_log KILL_LOG
                        Where to log killed queries to, granting permissions
                        to write to this file. (default:
                        /var/log/killed_queries.log)
  -klo, --kill_log_only
                        Only log the queries that would be killed, do not
                        actually kill them (default: False)

贡献

欢迎提出建议和贡献。请 fork 我并将 PR 创建回开发分支。