df134df901
Enabled by a new > 0 integer config value, "servers_per_port" in the [DEFAULT] config section for object-server and/or replication server configs. The setting's integer value determines how many different object-server workers handle requests for any single unique local port in the ring. In this mode, the parent swift-object-server process continues to run as the original user (i.e. root if low-port binding is required), binds to all ports as defined in the ring, and forks off the specified number of workers per listen socket. The child, per-port servers drop privileges and behave pretty much how object-server workers always have, except that because the ring has unique ports per disk, the object-servers will only be handling requests for a single disk. The parent process detects dead servers and restarts them (with the correct listen socket), starts missing servers when an updated ring file is found with a device on the server with a new port, and kills extraneous servers when their port is found to no longer be in the ring. The ring files are stat'ed at most every "ring_check_interval" seconds, as configured in the object-server config (same default of 15s). Immediately stopping all swift-object-worker processes still works by sending the parent a SIGTERM. Likewise, a SIGHUP to the parent process still causes the parent process to close all listen sockets and exit, allowing existing children to finish serving their existing requests. The drop_privileges helper function now has an optional param to suppress the setsid() call, which otherwise screws up the child workers' process management. The class method RingData.load() can be told to only load the ring metadata (i.e. everything except replica2part2dev_id) with the optional kwarg, header_only=True. This is used to keep the parent and all forked off workers from unnecessarily having full copies of all storage policy rings in memory. A new helper class, swift.common.storage_policy.BindPortsCache, provides a method to return a set of all device ports in all rings for the server on which it is instantiated (identified by its set of IP addresses). The BindPortsCache instance will track mtimes of ring files, so they are not opened more frequently than necessary. This patch includes enhancements to the probe tests and object-replicator/object-reconstructor config plumbing to allow the probe tests to work correctly both in the "normal" config (same IP but unique ports for each SAIO "server") and a server-per-port setup where each SAIO "server" must have a unique IP address and unique port per disk within each "server". The main probe tests only work with 4 servers and 4 disks, but you can see the difference in the rings for the EC probe tests where there are 2 disks per server for a total of 8 disks. Specifically, swift.common.ring.utils.is_local_device() will ignore the ports when the "my_port" argument is None. Then, object-replicator and object-reconstructor both set self.bind_port to None if server_per_port is enabled. Bonus improvement for IPv6 addresses in is_local_device(). This PR for vagrant-swift-all-in-one will aid in testing this patch: https://github.com/swiftstack/vagrant-swift-all-in-one/pull/16/ Also allow SAIO to answer is_local_device() better; common SAIO setups have multiple "servers" all on the same host with different ports for the different "servers" (which happen to match the IPs specified in the rings for the devices on each of those "servers"). However, you can configure the SAIO to have different localhost IP addresses (e.g. 127.0.0.1, 127.0.0.2, etc.) in the ring and in the servers' config files' bind_ip setting. This new whataremyips() implementation combined with a little plumbing allows is_local_device() to accurately answer, even on an SAIO. In the default case (an unspecified bind_ip defaults to '0.0.0.0') as well as an explict "bind to everything" like '0.0.0.0' or '::', whataremyips() behaves as it always has, returning all IP addresses for the server. Also updated probe tests to handle each "server" in the SAIO having a unique IP address. For some (noisy) benchmarks that show servers_per_port=X is at least as good as the same number of "normal" workers: https://gist.github.com/dbishop/c214f89ca708a6b1624a#file-summary-md Benchmarks showing the benefits of I/O isolation with a small number of slow disks: https://gist.github.com/dbishop/fd0ab067babdecfb07ca#file-results-md If you were wondering what the overhead of threads_per_disk looks like: https://gist.github.com/dbishop/1d14755fedc86a161718#file-tabular_results-md DocImpact Change-Id: I2239a4000b41a7e7cc53465ce794af49d44796c6
184 lines
7.2 KiB
Python
Executable File
184 lines
7.2 KiB
Python
Executable File
#!/usr/bin/python -u
|
|
# Copyright (c) 2010-2012 OpenStack Foundation
|
|
#
|
|
# Licensed under the Apache License, Version 2.0 (the "License");
|
|
# you may not use this file except in compliance with the License.
|
|
# You may obtain a copy of the License at
|
|
#
|
|
# http://www.apache.org/licenses/LICENSE-2.0
|
|
#
|
|
# Unless required by applicable law or agreed to in writing, software
|
|
# distributed under the License is distributed on an "AS IS" BASIS,
|
|
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
|
|
# implied.
|
|
# See the License for the specific language governing permissions and
|
|
# limitations under the License.
|
|
|
|
from os import listdir
|
|
from os.path import join as path_join
|
|
from unittest import main
|
|
from uuid import uuid4
|
|
|
|
from eventlet import GreenPool, Timeout
|
|
import eventlet
|
|
from sqlite3 import connect
|
|
from swiftclient import client
|
|
|
|
from swift.common import direct_client
|
|
from swift.common.exceptions import ClientException
|
|
from swift.common.utils import hash_path, readconf
|
|
from test.probe.common import kill_nonprimary_server, \
|
|
kill_server, ReplProbeTest, start_server
|
|
|
|
eventlet.monkey_patch(all=False, socket=True)
|
|
|
|
|
|
def get_db_file_path(obj_dir):
|
|
files = sorted(listdir(obj_dir), reverse=True)
|
|
for filename in files:
|
|
if filename.endswith('db'):
|
|
return path_join(obj_dir, filename)
|
|
|
|
|
|
class TestContainerFailures(ReplProbeTest):
|
|
|
|
def test_one_node_fails(self):
|
|
# Create container1
|
|
container1 = 'container-%s' % uuid4()
|
|
cpart, cnodes = self.container_ring.get_nodes(self.account, container1)
|
|
client.put_container(self.url, self.token, container1)
|
|
|
|
# Kill container1 servers excepting two of the primaries
|
|
kill_nonprimary_server(cnodes, self.ipport2server, self.pids)
|
|
kill_server((cnodes[0]['ip'], cnodes[0]['port']),
|
|
self.ipport2server, self.pids)
|
|
|
|
# Delete container1
|
|
client.delete_container(self.url, self.token, container1)
|
|
|
|
# Restart other container1 primary server
|
|
start_server((cnodes[0]['ip'], cnodes[0]['port']),
|
|
self.ipport2server, self.pids)
|
|
|
|
# Create container1/object1 (allowed because at least server thinks the
|
|
# container exists)
|
|
client.put_object(self.url, self.token, container1, 'object1', '123')
|
|
|
|
# Get to a final state
|
|
self.get_to_final_state()
|
|
|
|
# Assert all container1 servers indicate container1 is alive and
|
|
# well with object1
|
|
for cnode in cnodes:
|
|
self.assertEquals(
|
|
[o['name'] for o in direct_client.direct_get_container(
|
|
cnode, cpart, self.account, container1)[1]],
|
|
['object1'])
|
|
|
|
# Assert account level also indicates container1 is alive and
|
|
# well with object1
|
|
headers, containers = client.get_account(self.url, self.token)
|
|
self.assertEquals(headers['x-account-container-count'], '1')
|
|
self.assertEquals(headers['x-account-object-count'], '1')
|
|
self.assertEquals(headers['x-account-bytes-used'], '3')
|
|
|
|
def test_two_nodes_fail(self):
|
|
# Create container1
|
|
container1 = 'container-%s' % uuid4()
|
|
cpart, cnodes = self.container_ring.get_nodes(self.account, container1)
|
|
client.put_container(self.url, self.token, container1)
|
|
|
|
# Kill container1 servers excepting one of the primaries
|
|
cnp_ipport = kill_nonprimary_server(cnodes, self.ipport2server,
|
|
self.pids)
|
|
kill_server((cnodes[0]['ip'], cnodes[0]['port']),
|
|
self.ipport2server, self.pids)
|
|
kill_server((cnodes[1]['ip'], cnodes[1]['port']),
|
|
self.ipport2server, self.pids)
|
|
|
|
# Delete container1 directly to the one primary still up
|
|
direct_client.direct_delete_container(cnodes[2], cpart, self.account,
|
|
container1)
|
|
|
|
# Restart other container1 servers
|
|
start_server((cnodes[0]['ip'], cnodes[0]['port']),
|
|
self.ipport2server, self.pids)
|
|
start_server((cnodes[1]['ip'], cnodes[1]['port']),
|
|
self.ipport2server, self.pids)
|
|
start_server(cnp_ipport, self.ipport2server, self.pids)
|
|
|
|
# Get to a final state
|
|
self.get_to_final_state()
|
|
|
|
# Assert all container1 servers indicate container1 is gone (happens
|
|
# because the one node that knew about the delete replicated to the
|
|
# others.)
|
|
for cnode in cnodes:
|
|
try:
|
|
direct_client.direct_get_container(cnode, cpart, self.account,
|
|
container1)
|
|
except ClientException as err:
|
|
self.assertEquals(err.http_status, 404)
|
|
else:
|
|
self.fail("Expected ClientException but didn't get it")
|
|
|
|
# Assert account level also indicates container1 is gone
|
|
headers, containers = client.get_account(self.url, self.token)
|
|
self.assertEquals(headers['x-account-container-count'], '0')
|
|
self.assertEquals(headers['x-account-object-count'], '0')
|
|
self.assertEquals(headers['x-account-bytes-used'], '0')
|
|
|
|
def _get_container_db_files(self, container):
|
|
opart, onodes = self.container_ring.get_nodes(self.account, container)
|
|
onode = onodes[0]
|
|
db_files = []
|
|
for onode in onodes:
|
|
node_id = (onode['port'] - 6000) / 10
|
|
device = onode['device']
|
|
hash_str = hash_path(self.account, container)
|
|
server_conf = readconf(self.configs['container-server'][node_id])
|
|
devices = server_conf['app:container-server']['devices']
|
|
obj_dir = '%s/%s/containers/%s/%s/%s/' % (devices,
|
|
device, opart,
|
|
hash_str[-3:], hash_str)
|
|
db_files.append(get_db_file_path(obj_dir))
|
|
|
|
return db_files
|
|
|
|
def test_locked_container_dbs(self):
|
|
|
|
def run_test(num_locks, catch_503):
|
|
container = 'container-%s' % uuid4()
|
|
client.put_container(self.url, self.token, container)
|
|
db_files = self._get_container_db_files(container)
|
|
db_conns = []
|
|
for i in range(num_locks):
|
|
db_conn = connect(db_files[i])
|
|
db_conn.execute('begin exclusive transaction')
|
|
db_conns.append(db_conn)
|
|
if catch_503:
|
|
try:
|
|
client.delete_container(self.url, self.token, container)
|
|
except client.ClientException as err:
|
|
self.assertEquals(err.http_status, 503)
|
|
else:
|
|
self.fail("Expected ClientException but didn't get it")
|
|
else:
|
|
client.delete_container(self.url, self.token, container)
|
|
|
|
pool = GreenPool()
|
|
try:
|
|
with Timeout(15):
|
|
pool.spawn(run_test, 1, False)
|
|
pool.spawn(run_test, 2, True)
|
|
pool.spawn(run_test, 3, True)
|
|
pool.waitall()
|
|
except Timeout as err:
|
|
raise Exception(
|
|
"The server did not return a 503 on container db locks, "
|
|
"it just hangs: %s" % err)
|
|
|
|
|
|
if __name__ == '__main__':
|
|
main()
|