Mini Shell
"""
The 'sysbench' module is used to analyze the
performance of the minions, right from the master!
It measures various system parameters such as
CPU, Memory, File I/O, Threads and Mutex.
"""
import re
import salt.utils.path
def __virtual__():
"""
loads the module, if only sysbench is installed
"""
# finding the path of the binary
if salt.utils.path.which("sysbench"):
return "sysbench"
return (
False,
"The sysbench execution module failed to load: the sysbench binary is not in"
" the path.",
)
def _parser(result):
"""
parses the output into a dictionary
"""
# regexes to match
_total_time = re.compile(r"total time:\s*(\d*.\d*s)")
_total_execution = re.compile(r"event execution:\s*(\d*.\d*s?)")
_min_response_time = re.compile(r"min:\s*(\d*.\d*ms)")
_max_response_time = re.compile(r"max:\s*(\d*.\d*ms)")
_avg_response_time = re.compile(r"avg:\s*(\d*.\d*ms)")
_per_response_time = re.compile(r"95 percentile:\s*(\d*.\d*ms)")
# extracting data
total_time = re.search(_total_time, result).group(1)
total_execution = re.search(_total_execution, result).group(1)
min_response_time = re.search(_min_response_time, result).group(1)
max_response_time = re.search(_max_response_time, result).group(1)
avg_response_time = re.search(_avg_response_time, result).group(1)
per_response_time = re.search(_per_response_time, result)
if per_response_time is not None:
per_response_time = per_response_time.group(1)
# returning the data as dictionary
return {
"total time": total_time,
"total execution time": total_execution,
"minimum response time": min_response_time,
"maximum response time": max_response_time,
"average response time": avg_response_time,
"95 percentile": per_response_time,
}
def cpu():
"""
Tests for the CPU performance of minions.
CLI Examples:
.. code-block:: bash
salt '*' sysbench.cpu
"""
# Test data
max_primes = [500, 1000, 2500, 5000]
# Initializing the test variables
test_command = "sysbench --test=cpu --cpu-max-prime={0} run"
result = None
ret_val = {}
# Test beings!
for primes in max_primes:
key = f"Prime numbers limit: {primes}"
run_command = test_command.format(primes)
result = __salt__["cmd.run"](run_command)
ret_val[key] = _parser(result)
return ret_val
def threads():
"""
This tests the performance of the processor's scheduler
CLI Example:
.. code-block:: bash
salt '*' sysbench.threads
"""
# Test data
thread_yields = [100, 200, 500, 1000]
thread_locks = [2, 4, 8, 16]
# Initializing the test variables
test_command = "sysbench --num-threads=64 --test=threads "
test_command += "--thread-yields={0} --thread-locks={1} run "
result = None
ret_val = {}
# Test begins!
for yields, locks in zip(thread_yields, thread_locks):
key = f"Yields: {yields} Locks: {locks}"
run_command = test_command.format(yields, locks)
result = __salt__["cmd.run"](run_command)
ret_val[key] = _parser(result)
return ret_val
def mutex():
"""
Tests the implementation of mutex
CLI Examples:
.. code-block:: bash
salt '*' sysbench.mutex
"""
# Test options and the values they take
# --mutex-num = [50,500,1000]
# --mutex-locks = [10000,25000,50000]
# --mutex-loops = [2500,5000,10000]
# Test data (Orthogonal test cases)
mutex_num = [50, 50, 50, 500, 500, 500, 1000, 1000, 1000]
locks = [10000, 25000, 50000, 10000, 25000, 50000, 10000, 25000, 50000]
mutex_locks = []
mutex_locks.extend(locks)
mutex_loops = [2500, 5000, 10000, 10000, 2500, 5000, 5000, 10000, 2500]
# Initializing the test variables
test_command = "sysbench --num-threads=250 --test=mutex "
test_command += "--mutex-num={0} --mutex-locks={1} --mutex-loops={2} run "
result = None
ret_val = {}
# Test begins!
for num, locks, loops in zip(mutex_num, mutex_locks, mutex_loops):
key = f"Mutex: {num} Locks: {locks} Loops: {loops}"
run_command = test_command.format(num, locks, loops)
result = __salt__["cmd.run"](run_command)
ret_val[key] = _parser(result)
return ret_val
def memory():
"""
This tests the memory for read and write operations.
CLI Examples:
.. code-block:: bash
salt '*' sysbench.memory
"""
# test defaults
# --memory-block-size = 10M
# --memory-total-size = 1G
# We test memory read / write against global / local scope of memory
# Test data
memory_oper = ["read", "write"]
memory_scope = ["local", "global"]
# Initializing the test variables
test_command = "sysbench --num-threads=64 --test=memory "
test_command += "--memory-oper={0} --memory-scope={1} "
test_command += "--memory-block-size=1K --memory-total-size=32G run "
result = None
ret_val = {}
# Test begins!
for oper in memory_oper:
for scope in memory_scope:
key = f"Operation: {oper} Scope: {scope}"
run_command = test_command.format(oper, scope)
result = __salt__["cmd.run"](run_command)
ret_val[key] = _parser(result)
return ret_val
def fileio():
"""
This tests for the file read and write operations
Various modes of operations are
* sequential write
* sequential rewrite
* sequential read
* random read
* random write
* random read and write
The test works with 32 files with each file being 1Gb in size
The test consumes a lot of time. Be patient!
CLI Examples:
.. code-block:: bash
salt '*' sysbench.fileio
"""
# Test data
test_modes = ["seqwr", "seqrewr", "seqrd", "rndrd", "rndwr", "rndrw"]
# Initializing the required variables
test_command = "sysbench --num-threads=16 --test=fileio "
test_command += "--file-num=32 --file-total-size=1G --file-test-mode={0} "
result = None
ret_val = {}
# Test begins!
for mode in test_modes:
key = f"Mode: {mode}"
# Prepare phase
run_command = (test_command + "prepare").format(mode)
__salt__["cmd.run"](run_command)
# Test phase
run_command = (test_command + "run").format(mode)
result = __salt__["cmd.run"](run_command)
ret_val[key] = _parser(result)
# Clean up phase
run_command = (test_command + "cleanup").format(mode)
__salt__["cmd.run"](run_command)
return ret_val
def ping():
return True
Zerion Mini Shell 1.0