def report_check_glusterd_op_version(ctx):
cmd1 = "gluster v get all cluster.op-version"
cmd2 = "gluster v get all cluster.max-op-version"
try:
out1 = command_output(cmd1)
out2 = command_output(cmd2)
version1 = out1.split("\n")[-1].split(" ")[-1].strip()
version2 = out1.split("\n")[-1].split(" ")[-1].strip()
if version1 != version2:
ctx.warning("op-version is not up to date")
else:
ctx.ok("Glusterd is running", uptime_sec=out.strip())
except CommandError as e:
ctx.notok("Failed to check op-version")
logging.warn(
ctx.lf("Failed to check op-version", error_code=e[0], error=e[1]))
def report_gluster_memory_usage(ctx):
cmd = ['pgrep','gluster']
try:
out = command_output(cmd)
for pid in out.strip().split("\n"):
mem_cmd = "ps -p {} -o %mem".format(pid)
mem_out = command_output(mem_cmd)
mem_percent = mem_out.split('\n')[1].split('.')[0].strip()
if int(mem_percent) >= gluster_mem_limit:
proc_cmd = "ps -p {} -o comm=".format(pid)
proc_out = command_output(proc_cmd)
ctx.notok("Memory used by gluster process is at alarming level", process_name=proc_out.strip(), percentage=mem_percent)
except CommandError as e:
ctx.notok("Failed to get memory usage of gluster processes")
logging.warn(ctx.lf("pgrep/ps command failed",
error_code=e[0],
error=e[1]))
def test_command_output(self):
"""
Run a linux command without sudo:
seq - prints numeric sequences
"""
input_command = ['seq', '5']
seq_result = utils.command_output(input_command)
seq_expected = "1\n2\n3\n4\n5\n"
self.assertEqual(seq_result, seq_expected)
def report_gfid__mismatch_dht(ctx):
cmd = "grep 'gfid differs' " + logfile + " | grep -v grep | wc -l"
try:
out = command_output(cmd)
if int(out.strip()) > 0:
ctx.error("gfid mismatch found", no_of_mismatches=out.strip())
else:
ctx.ok("no gfid mismatch")
except CommandError as e:
ctx.notok("Command failure")
logging.warn(ctx.lf("Command Failure", error_code=e[0], error=e[1]))
def report_check_glusterd_uptime(ctx):
cmd = "ps -C glusterd --no-header -o etimes"
try:
out = command_output(cmd)
if out.strip() < "86400":
ctx.warning("Glusterd uptime is less than 24 hours",
uptime_sec=out.strip())
else:
ctx.ok("Glusterd is running", uptime_sec=out.strip())
except CommandError as e:
ctx.notok("Glusterd is not running")
logging.warn(
ctx.lf("Glusterd is not running", error_code=e[0], error=e[1]))
def report_coredump(ctx):
cmd = "ulimit -c"
try:
out = command_output(cmd)
if out.strip() == "unlimited":
ctx.ok("The maximum size of core files created is set to unlimted.")
else:
ctx.notok("The maximum size of core files created is NOT set to unlimited.")
except CommandError as e:
ctx.notok("ulimit check failed")
logging.warn(ctx.lf("ulimit check failed",
error_code=e[0],
error=e[1]))
def report_peer_disconnect(ctx):
cmd = "gluster peer status"
try:
out = command_output(cmd)
peer_count = out.split("\n")[0].split(":")[1].strip()
peer_conn_count = out.count("(Connected)")
dis_count = int(peer_count) - int(peer_conn_count)
if 0 < dis_count:
ctx.notok("One or more peer/s in disconnected/rejected state",
total_peer_count=int(peer_count),
disconnected_rejected_count=dis_count)
logging.warning("Peer status: \n" + out)
else:
ctx.ok("All peers are in connected state",
total_peer_count=int(peer_count),
connected_count=int(peer_conn_count))
except CommandError as e:
ctx.notok("Failed to check peer status")
logging.warn(
ctx.lf("peer status command failed", error_code=e[0], error=e[1]))
def report_system_memory_usage(ctx):
cmd = "free -m"
try:
out = command_output(cmd)
for line in out.split("\n"):
if "Mem" in line:
memtype, total, used, free, shared, cache, available = \
line.split()
elif "Swap" in line:
memtype, total, used, free = line.split()
else:
continue
percent = int(100 * float(used)/float(total))
if percent >= mem_used_limit:
ctx.notok("Memory used percentage on system is at alarming level", memtype=memtype.strip(':'), percentage=str(percent))
except CommandError as e:
ctx.notok("Failed to get memory usage")
logging.warn(ctx.lf("free command failed",
error_code=e[0],
error=e[1]))