From nobody Wed Nov 12 07:20:40 2025 Delivered-To: importer@patchew.org Received-SPF: pass (zoho.com: domain of gnu.org designates 209.51.188.17 as permitted sender) client-ip=209.51.188.17; envelope-from=qemu-devel-bounces+importer=patchew.org@nongnu.org; helo=lists.gnu.org; Authentication-Results: mx.zohomail.com; spf=pass (zoho.com: domain of gnu.org designates 209.51.188.17 as permitted sender) smtp.mailfrom=qemu-devel-bounces+importer=patchew.org@nongnu.org; dmarc=fail(p=none dis=none) header.from=virtuozzo.com ARC-Seal: i=1; a=rsa-sha256; t=1568740834; cv=none; d=zoho.com; s=zohoarc; b=Zj40UdnMJjasa4okQ1szlA1vmnqNSTqqgKFmzE7j57OgdsUCErdGth1yI2FgRRMbH8mDQ0SbJU0oHUYmrzXa0PPuZNDaQHuYqFN8Rbjlnaqd70Eh2faOPap2zH+/K4FGPLTd7nb8oBFE+BUheQDIXn4pnXJwifbN40o4jlhK5qs= ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=zoho.com; s=zohoarc; t=1568740834; h=Content-Transfer-Encoding:Cc:Date:From:In-Reply-To:List-Subscribe:List-Post:List-Id:List-Archive:List-Help:List-Unsubscribe:MIME-Version:Message-ID:References:Sender:Subject:To:ARC-Authentication-Results; bh=DaZCIEqPqvvkquCNLiiWVrfv9y0ML/R43aQrzZqTVtk=; b=jA99DuLTaiEU2omHDAYOn9309wQbuokd4IWjqqTHbFZqRBfw90wzgEJqJmaYLe+fdvF8lZ8C8GhvlhC7Iq2rsiSSaiavzI7RN/ybI0zRNsSl1dq9+h+qi+stp788UGLj0Mwh0PWOGKSpC6kiOFNsBZ7vnP3uHqrzRIrquUSwWGM= ARC-Authentication-Results: i=1; mx.zoho.com; spf=pass (zoho.com: domain of gnu.org designates 209.51.188.17 as permitted sender) smtp.mailfrom=qemu-devel-bounces+importer=patchew.org@nongnu.org; dmarc=fail header.from= (p=none dis=none) header.from= Return-Path: Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) by mx.zohomail.com with SMTPS id 1568740834193727.9433423189362; Tue, 17 Sep 2019 10:20:34 -0700 (PDT) Received: from localhost ([::1]:48822 helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1iAH9V-00049e-0M for importer@patchew.org; Tue, 17 Sep 2019 13:20:33 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]:50685) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1iAH2k-0006sr-KF for qemu-devel@nongnu.org; Tue, 17 Sep 2019 13:13:36 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1iAH2i-00036O-4m for qemu-devel@nongnu.org; Tue, 17 Sep 2019 13:13:34 -0400 Received: from relay.sw.ru ([185.231.240.75]:55312) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1iAH2c-000302-ND; Tue, 17 Sep 2019 13:13:27 -0400 Received: from [10.94.3.0] (helo=kvm.qa.sw.ru) by relay.sw.ru with esmtp (Exim 4.92.2) (envelope-from ) id 1iAH2Z-0003rq-RC; Tue, 17 Sep 2019 20:13:23 +0300 From: Vladimir Sementsov-Ogievskiy To: qemu-block@nongnu.org Date: Tue, 17 Sep 2019 20:13:21 +0300 Message-Id: <20190917171322.4127-3-vsementsov@virtuozzo.com> X-Mailer: git-send-email 2.21.0 In-Reply-To: <20190917171322.4127-1-vsementsov@virtuozzo.com> References: <20190917171322.4127-1-vsementsov@virtuozzo.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-detected-operating-system: by eggs.gnu.org: GNU/Linux 3.x X-Received-From: 185.231.240.75 Subject: [Qemu-devel] [PATCH v9 2/3] block/nbd: nbd reconnect X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: kwolf@redhat.com, vsementsov@virtuozzo.com, qemu-devel@nongnu.org, mreitz@redhat.com, stefanha@redhat.com, den@openvz.org Errors-To: qemu-devel-bounces+importer=patchew.org@nongnu.org Sender: "Qemu-devel" Content-Type: text/plain; charset="utf-8" Implement reconnect. To achieve this: 1. add new modes: connecting-wait: means, that reconnecting is in progress, and there were small number of reconnect attempts, so all requests are waiting for the connection. connecting-nowait: reconnecting is in progress, there were a lot of attempts of reconnect, all requests will return errors. two old modes are used too: connected: normal state quit: exiting after fatal error or on close Possible transitions are: * -> quit connecting-* -> connected connecting-wait -> connecting-nowait (transition is done after reconnect-delay seconds in connecting-wait mode) connected -> connecting-wait 2. Implement reconnect in connection_co. So, in connecting-* mode, connection_co, tries to reconnect unlimited times. 3. Retry nbd queries on channel error, if we are in connecting-wait state. Signed-off-by: Vladimir Sementsov-Ogievskiy --- block/nbd.c | 332 ++++++++++++++++++++++++++++++++++++++++++---------- 1 file changed, 269 insertions(+), 63 deletions(-) diff --git a/block/nbd.c b/block/nbd.c index 813c40d8f0..bec01d85c7 100644 --- a/block/nbd.c +++ b/block/nbd.c @@ -1,6 +1,7 @@ /* * QEMU Block driver for NBD * + * Copyright (c) 2019 Virtuozzo International GmbH. * Copyright (C) 2016 Red Hat, Inc. * Copyright (C) 2008 Bull S.A.S. * Author: Laurent Vivier @@ -55,6 +56,8 @@ typedef struct { } NBDClientRequest; =20 typedef enum NBDClientState { + NBD_CLIENT_CONNECTING_WAIT, + NBD_CLIENT_CONNECTING_NOWAIT, NBD_CLIENT_CONNECTED, NBD_CLIENT_QUIT } NBDClientState; @@ -67,8 +70,14 @@ typedef struct BDRVNBDState { CoMutex send_mutex; CoQueue free_sema; Coroutine *connection_co; + QemuCoSleepState *connection_co_sleep_ns_state; + bool drained; + bool wait_drained_end; int in_flight; NBDClientState state; + int connect_status; + Error *connect_err; + bool wait_in_flight; =20 NBDClientRequest requests[MAX_NBD_REQUESTS]; NBDReply reply; @@ -83,10 +92,21 @@ typedef struct BDRVNBDState { char *x_dirty_bitmap; } BDRVNBDState; =20 -/* @ret will be used for reconnect in future */ +static int nbd_client_connect(BlockDriverState *bs, Error **errp); + static void nbd_channel_error(BDRVNBDState *s, int ret) { - s->state =3D NBD_CLIENT_QUIT; + if (ret =3D=3D -EIO) { + if (s->state =3D=3D NBD_CLIENT_CONNECTED) { + s->state =3D s->reconnect_delay ? NBD_CLIENT_CONNECTING_WAIT : + NBD_CLIENT_CONNECTING_NOWAIT; + } + } else { + if (s->state =3D=3D NBD_CLIENT_CONNECTED) { + qio_channel_shutdown(s->ioc, QIO_CHANNEL_SHUTDOWN_BOTH, NULL); + } + s->state =3D NBD_CLIENT_QUIT; + } } =20 static void nbd_recv_coroutines_wake_all(BDRVNBDState *s) @@ -129,7 +149,13 @@ static void nbd_client_attach_aio_context(BlockDriverS= tate *bs, { BDRVNBDState *s =3D (BDRVNBDState *)bs->opaque; =20 - qio_channel_attach_aio_context(QIO_CHANNEL(s->ioc), new_context); + /* + * s->connection_co is either yielded from nbd_receive_reply or from + * nbd_reconnect_loop() + */ + if (s->state =3D=3D NBD_CLIENT_CONNECTED) { + qio_channel_attach_aio_context(QIO_CHANNEL(s->ioc), new_context); + } =20 bdrv_inc_in_flight(bs); =20 @@ -140,24 +166,151 @@ static void nbd_client_attach_aio_context(BlockDrive= rState *bs, aio_wait_bh_oneshot(new_context, nbd_client_attach_aio_context_bh, bs); } =20 +static void coroutine_fn nbd_client_co_drain_begin(BlockDriverState *bs) +{ + BDRVNBDState *s =3D (BDRVNBDState *)bs->opaque; =20 -static void nbd_teardown_connection(BlockDriverState *bs) + s->drained =3D true; + if (s->connection_co_sleep_ns_state) { + qemu_co_sleep_wake(s->connection_co_sleep_ns_state); + } +} + +static void coroutine_fn nbd_client_co_drain_end(BlockDriverState *bs) { BDRVNBDState *s =3D (BDRVNBDState *)bs->opaque; =20 - assert(s->ioc); + s->drained =3D false; + if (s->wait_drained_end) { + s->wait_drained_end =3D false; + aio_co_wake(s->connection_co); + } +} + =20 - /* finish any pending coroutines */ - qio_channel_shutdown(s->ioc, - QIO_CHANNEL_SHUTDOWN_BOTH, - NULL); +static void nbd_teardown_connection(BlockDriverState *bs) +{ + BDRVNBDState *s =3D (BDRVNBDState *)bs->opaque; + + if (s->state =3D=3D NBD_CLIENT_CONNECTED) { + /* finish any pending coroutines */ + assert(s->ioc); + qio_channel_shutdown(s->ioc, QIO_CHANNEL_SHUTDOWN_BOTH, NULL); + } + s->state =3D NBD_CLIENT_QUIT; + if (s->connection_co) { + if (s->connection_co_sleep_ns_state) { + qemu_co_sleep_wake(s->connection_co_sleep_ns_state); + } + } BDRV_POLL_WHILE(bs, s->connection_co); +} =20 - nbd_client_detach_aio_context(bs); - object_unref(OBJECT(s->sioc)); - s->sioc =3D NULL; - object_unref(OBJECT(s->ioc)); - s->ioc =3D NULL; +static bool nbd_client_connecting(BDRVNBDState *s) +{ + return s->state =3D=3D NBD_CLIENT_CONNECTING_WAIT || + s->state =3D=3D NBD_CLIENT_CONNECTING_NOWAIT; +} + +static bool nbd_client_connecting_wait(BDRVNBDState *s) +{ + return s->state =3D=3D NBD_CLIENT_CONNECTING_WAIT; +} + +static coroutine_fn void nbd_reconnect_attempt(BDRVNBDState *s) +{ + Error *local_err =3D NULL; + + if (!nbd_client_connecting(s)) { + return; + } + + /* Wait for completion of all in-flight requests */ + + qemu_co_mutex_lock(&s->send_mutex); + + while (s->in_flight > 0) { + qemu_co_mutex_unlock(&s->send_mutex); + nbd_recv_coroutines_wake_all(s); + s->wait_in_flight =3D true; + qemu_coroutine_yield(); + s->wait_in_flight =3D false; + qemu_co_mutex_lock(&s->send_mutex); + } + + qemu_co_mutex_unlock(&s->send_mutex); + + if (!nbd_client_connecting(s)) { + return; + } + + /* + * Now we are sure that nobody is accessing the channel, and no one wi= ll + * try until we set the state to CONNECTED. + */ + + /* Finalize previous connection if any */ + if (s->ioc) { + nbd_client_detach_aio_context(s->bs); + object_unref(OBJECT(s->sioc)); + s->sioc =3D NULL; + object_unref(OBJECT(s->ioc)); + s->ioc =3D NULL; + } + + s->connect_status =3D nbd_client_connect(s->bs, &local_err); + error_free(s->connect_err); + s->connect_err =3D NULL; + error_propagate(&s->connect_err, local_err); + local_err =3D NULL; + + if (s->connect_status < 0) { + /* failed attempt */ + return; + } + + /* successfully connected */ + s->state =3D NBD_CLIENT_CONNECTED; + qemu_co_queue_restart_all(&s->free_sema); +} + +static coroutine_fn void nbd_reconnect_loop(BDRVNBDState *s) +{ + uint64_t start_time_ns =3D qemu_clock_get_ns(QEMU_CLOCK_REALTIME); + uint64_t delay_ns =3D s->reconnect_delay * NANOSECONDS_PER_SECOND; + uint64_t timeout =3D 1 * NANOSECONDS_PER_SECOND; + uint64_t max_timeout =3D 16 * NANOSECONDS_PER_SECOND; + + nbd_reconnect_attempt(s); + + while (nbd_client_connecting(s)) { + if (s->state =3D=3D NBD_CLIENT_CONNECTING_WAIT && + qemu_clock_get_ns(QEMU_CLOCK_REALTIME) - start_time_ns > delay= _ns) + { + s->state =3D NBD_CLIENT_CONNECTING_NOWAIT; + qemu_co_queue_restart_all(&s->free_sema); + } + + qemu_co_sleep_ns_wakeable(QEMU_CLOCK_REALTIME, timeout, + &s->connection_co_sleep_ns_state); + if (s->drained) { + bdrv_dec_in_flight(s->bs); + s->wait_drained_end =3D true; + while (s->drained) { + /* + * We may be entered once from nbd_client_attach_aio_conte= xt_bh + * and then from nbd_client_co_drain_end. So here is a loo= p. + */ + qemu_coroutine_yield(); + } + bdrv_inc_in_flight(s->bs); + } + if (timeout < max_timeout) { + timeout *=3D 2; + } + + nbd_reconnect_attempt(s); + } } =20 static coroutine_fn void nbd_connection_entry(void *opaque) @@ -177,16 +330,26 @@ static coroutine_fn void nbd_connection_entry(void *o= paque) * Therefore we keep an additional in_flight reference all the tim= e and * only drop it temporarily here. */ + + if (nbd_client_connecting(s)) { + nbd_reconnect_loop(s); + } + + if (s->state !=3D NBD_CLIENT_CONNECTED) { + continue; + } + assert(s->reply.handle =3D=3D 0); ret =3D nbd_receive_reply(s->bs, s->ioc, &s->reply, &local_err); =20 if (local_err) { trace_nbd_read_reply_entry_fail(ret, error_get_pretty(local_er= r)); error_free(local_err); + local_err =3D NULL; } if (ret <=3D 0) { nbd_channel_error(s, ret ? ret : -EIO); - break; + continue; } =20 /* @@ -201,7 +364,7 @@ static coroutine_fn void nbd_connection_entry(void *opa= que) (nbd_reply_is_structured(&s->reply) && !s->info.structured_rep= ly)) { nbd_channel_error(s, -EINVAL); - break; + continue; } =20 /* @@ -220,10 +383,19 @@ static coroutine_fn void nbd_connection_entry(void *o= paque) qemu_coroutine_yield(); } =20 + qemu_co_queue_restart_all(&s->free_sema); nbd_recv_coroutines_wake_all(s); bdrv_dec_in_flight(s->bs); =20 s->connection_co =3D NULL; + if (s->ioc) { + nbd_client_detach_aio_context(s->bs); + object_unref(OBJECT(s->sioc)); + s->sioc =3D NULL; + object_unref(OBJECT(s->ioc)); + s->ioc =3D NULL; + } + aio_wait_kick(); } =20 @@ -235,7 +407,7 @@ static int nbd_co_send_request(BlockDriverState *bs, int rc, i =3D -1; =20 qemu_co_mutex_lock(&s->send_mutex); - while (s->in_flight =3D=3D MAX_NBD_REQUESTS) { + while (s->in_flight =3D=3D MAX_NBD_REQUESTS || nbd_client_connecting_w= ait(s)) { qemu_co_queue_wait(&s->free_sema, &s->send_mutex); } =20 @@ -286,7 +458,11 @@ err: s->requests[i].coroutine =3D NULL; s->in_flight--; } - qemu_co_queue_next(&s->free_sema); + if (s->in_flight =3D=3D 0 && s->wait_in_flight) { + aio_co_wake(s->connection_co); + } else { + qemu_co_queue_next(&s->free_sema); + } } qemu_co_mutex_unlock(&s->send_mutex); return rc; @@ -666,10 +842,15 @@ static coroutine_fn int nbd_co_receive_one_chunk( } else { /* For assert at loop start in nbd_connection_entry */ *reply =3D s->reply; - s->reply.handle =3D 0; } + s->reply.handle =3D 0; =20 - if (s->connection_co) { + if (s->connection_co && !s->wait_in_flight) { + /* + * We must check s->wait_in_flight, because we may entered by + * nbd_recv_coroutines_wake_all(), in this case we should not + * wake connection_co here, it will woken by last request. + */ aio_co_wake(s->connection_co); } =20 @@ -781,7 +962,11 @@ break_loop: =20 qemu_co_mutex_lock(&s->send_mutex); s->in_flight--; - qemu_co_queue_next(&s->free_sema); + if (s->in_flight =3D=3D 0 && s->wait_in_flight) { + aio_co_wake(s->connection_co); + } else { + qemu_co_queue_next(&s->free_sema); + } qemu_co_mutex_unlock(&s->send_mutex); =20 return false; @@ -927,20 +1112,26 @@ static int nbd_co_request(BlockDriverState *bs, NBDR= equest *request, } else { assert(request->type !=3D NBD_CMD_WRITE); } - ret =3D nbd_co_send_request(bs, request, write_qiov); - if (ret < 0) { - return ret; - } =20 - ret =3D nbd_co_receive_return_code(s, request->handle, - &request_ret, &local_err); - if (local_err) { - trace_nbd_co_request_fail(request->from, request->len, request->ha= ndle, - request->flags, request->type, - nbd_cmd_lookup(request->type), - ret, error_get_pretty(local_err)); - error_free(local_err); - } + do { + ret =3D nbd_co_send_request(bs, request, write_qiov); + if (ret < 0) { + continue; + } + + ret =3D nbd_co_receive_return_code(s, request->handle, + &request_ret, &local_err); + if (local_err) { + trace_nbd_co_request_fail(request->from, request->len, + request->handle, request->flags, + request->type, + nbd_cmd_lookup(request->type), + ret, error_get_pretty(local_err)); + error_free(local_err); + local_err =3D NULL; + } + } while (ret < 0 && nbd_client_connecting_wait(s)); + return ret ? ret : request_ret; } =20 @@ -981,20 +1172,24 @@ static int nbd_client_co_preadv(BlockDriverState *bs= , uint64_t offset, request.len -=3D slop; } =20 - ret =3D nbd_co_send_request(bs, &request, NULL); - if (ret < 0) { - return ret; - } + do { + ret =3D nbd_co_send_request(bs, &request, NULL); + if (ret < 0) { + continue; + } + + ret =3D nbd_co_receive_cmdread_reply(s, request.handle, offset, qi= ov, + &request_ret, &local_err); + if (local_err) { + trace_nbd_co_request_fail(request.from, request.len, request.h= andle, + request.flags, request.type, + nbd_cmd_lookup(request.type), + ret, error_get_pretty(local_err)); + error_free(local_err); + local_err =3D NULL; + } + } while (ret < 0 && nbd_client_connecting_wait(s)); =20 - ret =3D nbd_co_receive_cmdread_reply(s, request.handle, offset, qiov, - &request_ret, &local_err); - if (local_err) { - trace_nbd_co_request_fail(request.from, request.len, request.handl= e, - request.flags, request.type, - nbd_cmd_lookup(request.type), - ret, error_get_pretty(local_err)); - error_free(local_err); - } return ret ? ret : request_ret; } =20 @@ -1131,20 +1326,25 @@ static int coroutine_fn nbd_client_co_block_status( if (s->info.min_block) { assert(QEMU_IS_ALIGNED(request.len, s->info.min_block)); } - ret =3D nbd_co_send_request(bs, &request, NULL); - if (ret < 0) { - return ret; - } + do { + ret =3D nbd_co_send_request(bs, &request, NULL); + if (ret < 0) { + continue; + } + + ret =3D nbd_co_receive_blockstatus_reply(s, request.handle, bytes, + &extent, &request_ret, + &local_err); + if (local_err) { + trace_nbd_co_request_fail(request.from, request.len, request.h= andle, + request.flags, request.type, + nbd_cmd_lookup(request.type), + ret, error_get_pretty(local_err)); + error_free(local_err); + local_err =3D NULL; + } + } while (ret < 0 && nbd_client_connecting_wait(s)); =20 - ret =3D nbd_co_receive_blockstatus_reply(s, request.handle, bytes, - &extent, &request_ret, &local_e= rr); - if (local_err) { - trace_nbd_co_request_fail(request.from, request.len, request.handl= e, - request.flags, request.type, - nbd_cmd_lookup(request.type), - ret, error_get_pretty(local_err)); - error_free(local_err); - } if (ret < 0 || request_ret < 0) { return ret ? ret : request_ret; } @@ -1163,9 +1363,9 @@ static void nbd_client_close(BlockDriverState *bs) BDRVNBDState *s =3D (BDRVNBDState *)bs->opaque; NBDRequest request =3D { .type =3D NBD_CMD_DISC }; =20 - assert(s->ioc); - - nbd_send_request(s->ioc, &request); + if (s->ioc) { + nbd_send_request(s->ioc, &request); + } =20 nbd_teardown_connection(bs); } @@ -1808,6 +2008,8 @@ static BlockDriver bdrv_nbd =3D { .bdrv_getlength =3D nbd_getlength, .bdrv_detach_aio_context =3D nbd_client_detach_aio_context, .bdrv_attach_aio_context =3D nbd_client_attach_aio_context, + .bdrv_co_drain_begin =3D nbd_client_co_drain_begin, + .bdrv_co_drain_end =3D nbd_client_co_drain_end, .bdrv_refresh_filename =3D nbd_refresh_filename, .bdrv_co_block_status =3D nbd_client_co_block_status, .bdrv_dirname =3D nbd_dirname, @@ -1830,6 +2032,8 @@ static BlockDriver bdrv_nbd_tcp =3D { .bdrv_getlength =3D nbd_getlength, .bdrv_detach_aio_context =3D nbd_client_detach_aio_context, .bdrv_attach_aio_context =3D nbd_client_attach_aio_context, + .bdrv_co_drain_begin =3D nbd_client_co_drain_begin, + .bdrv_co_drain_end =3D nbd_client_co_drain_end, .bdrv_refresh_filename =3D nbd_refresh_filename, .bdrv_co_block_status =3D nbd_client_co_block_status, .bdrv_dirname =3D nbd_dirname, @@ -1852,6 +2056,8 @@ static BlockDriver bdrv_nbd_unix =3D { .bdrv_getlength =3D nbd_getlength, .bdrv_detach_aio_context =3D nbd_client_detach_aio_context, .bdrv_attach_aio_context =3D nbd_client_attach_aio_context, + .bdrv_co_drain_begin =3D nbd_client_co_drain_begin, + .bdrv_co_drain_end =3D nbd_client_co_drain_end, .bdrv_refresh_filename =3D nbd_refresh_filename, .bdrv_co_block_status =3D nbd_client_co_block_status, .bdrv_dirname =3D nbd_dirname, --=20 2.21.0