From patchwork Thu Mar 14 21:12:07 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Scott Mayhew X-Patchwork-Id: 10853667 Return-Path: Received: from mail.wl.linuxfoundation.org (pdx-wl-mail.web.codeaurora.org [172.30.200.125]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 77BD117E6 for ; Thu, 14 Mar 2019 21:12:15 +0000 (UTC) Received: from mail.wl.linuxfoundation.org (localhost [127.0.0.1]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id 649762A69B for ; Thu, 14 Mar 2019 21:12:15 +0000 (UTC) Received: by mail.wl.linuxfoundation.org (Postfix, from userid 486) id 58F9F2A6BB; Thu, 14 Mar 2019 21:12:15 +0000 (UTC) X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on pdx-wl-mail.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-7.9 required=2.0 tests=BAYES_00,MAILING_LIST_MULTI, RCVD_IN_DNSWL_HI autolearn=ham version=3.3.1 Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id 341D02A6FE for ; Thu, 14 Mar 2019 21:12:13 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727412AbfCNVML (ORCPT ); Thu, 14 Mar 2019 17:12:11 -0400 Received: from mx1.redhat.com ([209.132.183.28]:33640 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727715AbfCNVML (ORCPT ); Thu, 14 Mar 2019 17:12:11 -0400 Received: from smtp.corp.redhat.com (int-mx04.intmail.prod.int.phx2.redhat.com [10.5.11.14]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id 723A730EBB51; Thu, 14 Mar 2019 21:12:11 +0000 (UTC) Received: from coeurl.usersys.redhat.com (ovpn-122-19.rdu2.redhat.com [10.10.122.19]) by smtp.corp.redhat.com (Postfix) with ESMTP id 2A9C05D9C4; Thu, 14 Mar 2019 21:12:10 +0000 (UTC) Received: by coeurl.usersys.redhat.com (Postfix, from userid 1000) id 7FDD3213F7; Thu, 14 Mar 2019 17:12:10 -0400 (EDT) From: Scott Mayhew To: bfields@fieldses.org Cc: jlayton@kernel.org, linux-nfs@vger.kernel.org Subject: [pynfs PATCH 1/4] nfs4.1: add some reboot tests Date: Thu, 14 Mar 2019 17:12:07 -0400 Message-Id: <20190314211210.7454-2-smayhew@redhat.com> In-Reply-To: <20190314211210.7454-1-smayhew@redhat.com> References: <20190314211210.7454-1-smayhew@redhat.com> X-Scanned-By: MIMEDefang 2.79 on 10.5.11.14 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.5.110.49]); Thu, 14 Mar 2019 21:12:11 +0000 (UTC) Sender: linux-nfs-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-nfs@vger.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP REBT2a, REBT2b, and REBT2c test recovery with multiple clients following a server reboot. Signed-off-by: Scott Mayhew --- nfs4.1/server41tests/st_reboot.py | 151 +++++++++++++++++++++++++++++- 1 file changed, 147 insertions(+), 4 deletions(-) diff --git a/nfs4.1/server41tests/st_reboot.py b/nfs4.1/server41tests/st_reboot.py index b852ded..0216127 100644 --- a/nfs4.1/server41tests/st_reboot.py +++ b/nfs4.1/server41tests/st_reboot.py @@ -1,25 +1,33 @@ from xdrdef.nfs4_const import * from xdrdef.nfs4_type import * -from .environment import check, fail, create_file, open_file, create_confirm +from .environment import check, fail, create_file, open_file, create_confirm, close_file import sys import os -import nfs4lib +import time +import logging import nfs_ops op = nfs_ops.NFS4ops() from rpc import RPCTimeout # NOTE - reboot tests are NOT part of the standard test suite +log = logging.getLogger("test.env") + def _getleasetime(sess): res = sess.compound([op.putrootfh(), op.getattr(1 << FATTR4_LEASE_TIME)]) return res.resarray[-1].obj_attributes[FATTR4_LEASE_TIME] -def _waitForReboot(c, sess, env): +def _waitForReboot(env): env.serverhelper("reboot") # Wait until the server is back up. # The following blocks until it gets a response, # which happens when the server comes back up. env.c1.c1 = env.c1.connect(env.c1.server_address) + # Go ahead and whack the cached session and client ids now + # to avoid errors in Environment.finish(). + env.c1.sessions = {} + env.c1.clients = {} + return int(time.time()) def create_session(c, cred=None, flags=0): """Send a simple CREATE_SESSION""" @@ -51,7 +59,7 @@ def testRebootValid(t, env): reclaim_complete(sess) fh, stateid = create_confirm(sess, owner) sleeptime = 5 + _getleasetime(sess) - _waitForReboot(c, sess, env) + _waitForReboot(env) try: res = create_session(c) check(res, NFS4ERR_STALE_CLIENTID, "Reclaim using old clientid") @@ -67,3 +75,138 @@ def testRebootValid(t, env): reclaim_complete(sess) finally: env.sleep(sleeptime, "Waiting for grace period to end") + +class State(object): + def __init__(self, name, owner, c, sess, fh): + self.name = name + self.owner = owner + self.c = c + self.sess = sess + self.fh = fh + +def doTestOneClientGrace(t, env, state): + res = state.sess.compound([]) + check(res, NFS4ERR_BADSESSION, "Bare sequence after reboot") + res = create_session(state.c) + check(res, NFS4ERR_STALE_CLIENTID, "Reclaim using old clientid") + c = env.c1.new_client(state.name) + state.c = c + sess = c.create_session() + state.sess = sess + lease_time = _getleasetime(sess) + res = open_file(sess, state.owner, path=state.fh, + claim_type=CLAIM_PREVIOUS, + access=OPEN4_SHARE_ACCESS_BOTH, + deny=OPEN4_SHARE_DENY_NONE, + deleg_type=OPEN_DELEGATE_NONE) + check(res, msg="Reclaim using newly created clientid") + fh = res.resarray[-1].object + stateid = res.resarray[-2].stateid + reclaim_complete(sess) + close_file(sess, fh, stateid=stateid) + res = open_file(sess, state.owner, claim_type=CLAIM_NULL, + access=OPEN4_SHARE_ACCESS_BOTH, + deny=OPEN4_SHARE_DENY_NONE, + deleg_type=OPEN_DELEGATE_NONE) + check(res, NFS4ERR_GRACE, "New open during grace") + return lease_time + +def doTestOneClientNoGrace(t, env, state): + res = open_file(state.sess, state.owner, claim_type=CLAIM_NULL, + access=OPEN4_SHARE_ACCESS_BOTH, + deny=OPEN4_SHARE_DENY_NONE, + deleg_type=OPEN_DELEGATE_NONE) + if (res.status == NFS4ERR_GRACE): + return res + check(res, msg="New open after all clients done reclaiming") + fh = res.resarray[-1].object + stateid = res.resarray[-2].stateid + close_file(state.sess, fh, stateid=stateid) + return res + +# The server may have lifted the grace period early, but it's not obligated. +# Keep looping until all the clients have done a normal open. If the server +# didn't lift the grace period early we don't want to fail the test, but we +# do want to log a message. +def doTestAllClientsNoGrace(t, env, states): + all_done = False + warn_grace = False + start_time = int(time.time()) + ok_time = 0 + while not all_done: + all_done = True + for state in states: + res = doTestOneClientNoGrace(t, env, state) + if res.status == NFS4ERR_GRACE: + warn_grace = True + all_done = False + elif not ok_time: + ok_time = int(time.time()) + if not all_done: + time.sleep(1) + if warn_grace: + lift_time = ok_time - start_time + log.warn("server took approximately %d seconds to lift grace " + "after all clients reclaimed" % lift_time) + +def doTestRebootWithNClients(t, env, n=10): + boot_time = int(time.time()) + lease_time = 90 + states = [] + block = env.c1.new_client_session("%s_block" % env.testname(t)) + for i in range(n): + name = "%s_client_%i" % (env.testname(t), i) + owner = "owner_%s" % name + c = env.c1.new_client(name) + sess = c.create_session() + reclaim_complete(sess) + fh, stateid = create_confirm(sess, owner) + states.append(State(name, owner, c, sess, fh)) + lease_time = _getleasetime(sess) + boot_time = _waitForReboot(env) + + try: + for i in range(n): + lease_time = doTestOneClientGrace(t, env, states[i]) + + # At this point, all clients should have recovered except for 'block'. + # Recover that one now. + block = env.c1.new_client_session("%s_block" % env.testname(t)) + + # The server may have lifted the grace period early. Test it. + doTestAllClientsNoGrace(t, env, states) + + # If the test went normally, then the grace period should have already + # ended. If we got some unexpected error, then wait a bit for the server + # to expire the clients so cleanup can proceed more smoothly. + except: + grace_end_time = boot_time + lease_time + 5 + now = int(time.time()) + if now < grace_end_time: + sleep_time = grace_end_time - now + env.sleep(sleep_time, "Waiting for grace period to end") + raise + +def testRebootWithManyClients(t, env): + """Reboot with many clients + + FLAGS: reboot + CODE: REBT2a + """ + return doTestRebootWithNClients(t, env) + +def testRebootWithManyManyClients(t, env): + """Reboot with many many clients + + FLAGS: reboot + CODE: REBT2b + """ + return doTestRebootWithNClients(t, env, 100) + +def testRebootWithManyManyManyClients(t, env): + """Reboot with many many many clients + + FLAGS: reboot + CODE: REBT2c + """ + return doTestRebootWithNClients(t, env, 1000)