From: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
To: zhanghailiang <zhang.zhanghailiang@huawei.com>
Cc: lizhijian@cn.fujitsu.com, quintela@redhat.com,
yunhong.jiang@intel.com, eddie.dong@intel.com,
peter.huangpeng@huawei.com, qemu-devel@nongnu.org,
arei.gonglei@huawei.com, stefanha@redhat.com,
amit.shah@redhat.com
Subject: Re: [Qemu-devel] [PATCH COLO-Frame v10 09/38] COLO: Implement colo checkpoint protocol
Date: Fri, 6 Nov 2015 18:26:32 +0000 [thread overview]
Message-ID: <20151106182632.GH4267@work-vm> (raw)
In-Reply-To: <1446551816-15768-10-git-send-email-zhang.zhanghailiang@huawei.com>
* zhanghailiang (zhang.zhanghailiang@huawei.com) wrote:
> We need communications protocol of user-defined to control the checkpoint
> process.
>
> The new checkpoint request is started by Primary VM, and the interactive process
> like below:
> Checkpoint synchronizing points,
>
> Primary Secondary
> 'checkpoint-request' @ ----------------------------->
> Suspend (In hybrid mode)
> 'checkpoint-reply' <------------------------------ @
> Suspend&Save state
Why is this initial pair necessary? Can't you just start with the vmstate-send
and save the extra request pair/round trip? On the 2nd checkpoint we know
the SVM already received the previous checkpoint because we got it's first
vmstate-load.
I guess in full-COLO (rather than simple checkpoint) you can get the secondary to
do some of it's stopping/cleanup after it sends the checkpoint-reply
but before vmstate-send, so you can hide some of the time.
(Perhaps add a comment to explain)
> 'vmstate-send' @ ----------------------------->
> Send state Receive state
> 'vmstate-received' <------------------------------ @
> Release packets Load state
> 'vmstate-load' <------------------------------ @
> Resume Resume (In hybrid mode)
>
> Start Comparing (In hybrid mode)
> NOTE:
> 1) '@' who sends the message
> 2) Every sync-point is synchronized by two sides with only
> one handshake(single direction) for low-latency.
> If more strict synchronization is required, a opposite direction
> sync-point should be added.
> 3) Since sync-points are single direction, the remote side may
> go forward a lot when this side just receives the sync-point.
> 4) For now, we only support 'periodic' checkpoint, for which
> the Secondary VM is not running, later we will support 'hybrid' mode.
>
> Signed-off-by: zhanghailiang <zhang.zhanghailiang@huawei.com>
> Signed-off-by: Li Zhijian <lizhijian@cn.fujitsu.com>
> Signed-off-by: Gonglei <arei.gonglei@huawei.com>
> Cc: Eric Blake <eblake@redhat.com>
> ---
> v10:
> - Rename enum COLOCmd to COLOCommand (Eric's suggestion).
> - Remove unused 'ram-steal'
> ---
> migration/colo.c | 192 ++++++++++++++++++++++++++++++++++++++++++++++++++++++-
> qapi-schema.json | 27 ++++++++
> trace-events | 2 +
> 3 files changed, 219 insertions(+), 2 deletions(-)
>
> diff --git a/migration/colo.c b/migration/colo.c
> index 4fdf3a9..2510762 100644
> --- a/migration/colo.c
> +++ b/migration/colo.c
> @@ -10,10 +10,12 @@
> * later. See the COPYING file in the top-level directory.
> */
>
> +#include <unistd.h>
> #include "sysemu/sysemu.h"
> #include "migration/colo.h"
> #include "trace.h"
> #include "qemu/error-report.h"
> +#include "qemu/sockets.h"
>
> bool colo_supported(void)
> {
> @@ -34,6 +36,103 @@ bool migration_incoming_in_colo_state(void)
> return mis && (mis->state == MIGRATION_STATUS_COLO);
> }
>
> +/* colo checkpoint control helper */
> +static int colo_ctl_put(QEMUFile *f, uint32_t cmd, uint64_t value)
> +{
> + int ret = 0;
> +
> + qemu_put_be32(f, cmd);
> + qemu_put_be64(f, value);
> + qemu_fflush(f);
> +
> + ret = qemu_file_get_error(f);
> + trace_colo_ctl_put(COLOCommand_lookup[cmd], value);
> +
> + return ret;
> +}
> +
> +static int colo_ctl_get_cmd(QEMUFile *f, uint32_t *cmd)
> +{
> + int ret = 0;
> +
> + *cmd = qemu_get_be32(f);
> + ret = qemu_file_get_error(f);
> + if (ret < 0) {
> + return ret;
> + }
> + if (*cmd >= COLO_COMMAND_MAX) {
> + error_report("Invalid colo command, get cmd:%d", *cmd);
> + return -EINVAL;
> + }
> + trace_colo_ctl_get(COLOCommand_lookup[*cmd]);
> +
> + return 0;
> +}
> +
> +static int colo_ctl_get(QEMUFile *f, uint32_t require)
> +{
> + int ret;
> + uint32_t cmd;
> + uint64_t value;
> +
> + ret = colo_ctl_get_cmd(f, &cmd);
> + if (ret < 0) {
> + return ret;
> + }
> + if (cmd != require) {
> + error_report("Unexpect colo command, expect:%d, but get cmd:%d",
> + require, cmd);
> + return -EINVAL;
> + }
> +
> + value = qemu_get_be64(f);
> + ret = qemu_file_get_error(f);
> + if (ret < 0) {
> + return ret;
> + }
> +
> + return value;
> +}
Should the return type be uint64_t since you're returning value?
But then you're also using it to return an error code; so perhaps
it might be better to have a uint64_t *value parameter to
return the value separately; or define the range that the value
can actually take.
(Also very minor typo: 'got' not 'get' in a few errors)
> +static int colo_do_checkpoint_transaction(MigrationState *s)
> +{
> + int ret;
> +
> + ret = colo_ctl_put(s->to_dst_file, COLO_COMMAND_CHECKPOINT_REQUEST, 0);
> + if (ret < 0) {
> + goto out;
> + }
> +
> + ret = colo_ctl_get(s->from_dst_file, COLO_COMMAND_CHECKPOINT_REPLY);
> + if (ret < 0) {
> + goto out;
> + }
> +
> + /* TODO: suspend and save vm state to colo buffer */
> +
> + ret = colo_ctl_put(s->to_dst_file, COLO_COMMAND_VMSTATE_SEND, 0);
> + if (ret < 0) {
> + goto out;
> + }
> +
> + /* TODO: send vmstate to Secondary */
> +
> + ret = colo_ctl_get(s->from_dst_file, COLO_COMMAND_VMSTATE_RECEIVED);
> + if (ret < 0) {
> + goto out;
> + }> +
> + ret = colo_ctl_get(s->from_dst_file, COLO_COMMAND_VMSTATE_LOADED);
> + if (ret < 0) {
> + goto out;
> + }
> +
> + /* TODO: resume Primary */
> +
> +out:
> + return ret;
> +}
> +
> static void colo_process_checkpoint(MigrationState *s)
> {
> int fd, ret = 0;
> @@ -51,12 +150,27 @@ static void colo_process_checkpoint(MigrationState *s)
> goto out;
> }
>
> + /*
> + * Wait for Secondary finish loading vm states and enter COLO
> + * restore.
> + */
> + ret = colo_ctl_get(s->from_dst_file, COLO_COMMAND_CHECKPOINT_READY);
> + if (ret < 0) {
> + goto out;
> + }
> +
> qemu_mutex_lock_iothread();
> vm_start();
> qemu_mutex_unlock_iothread();
> trace_colo_vm_state_change("stop", "run");
>
> - /*TODO: COLO checkpoint savevm loop*/
> + while (s->state == MIGRATION_STATUS_COLO) {
> + /* start a colo checkpoint */
> + ret = colo_do_checkpoint_transaction(s);
> + if (ret < 0) {
> + goto out;
> + }
> + }
>
> out:
> if (ret < 0) {
> @@ -79,6 +193,39 @@ void migrate_start_colo_process(MigrationState *s)
> qemu_mutex_lock_iothread();
> }
>
> +/*
> + * return:
> + * 0: start a checkpoint
> + * -1: some error happened, exit colo restore
> + */
> +static int colo_wait_handle_cmd(QEMUFile *f, int *checkpoint_request)
> +{
> + int ret;
> + uint32_t cmd;
> + uint64_t value;
> +
> + ret = colo_ctl_get_cmd(f, &cmd);
> + if (ret < 0) {
> + /* do failover ? */
> + return ret;
> + }
> + /* Fix me: this value should be 0, which is not so good,
> + * should be used for checking ?
> + */
> + value = qemu_get_be64(f);
> + if (value != 0) {
should output error message as well?
> + return -EINVAL;
> + }
> +
> + switch (cmd) {
> + case COLO_COMMAND_CHECKPOINT_REQUEST:
> + *checkpoint_request = 1;
> + return 0;
> + default:
> + return -EINVAL;
> + }
> +}
> +
> void *colo_process_incoming_thread(void *opaque)
> {
> MigrationIncomingState *mis = opaque;
> @@ -98,7 +245,48 @@ void *colo_process_incoming_thread(void *opaque)
> error_report("colo incoming thread: Open QEMUFile to_src_file failed");
> goto out;
> }
> - /* TODO: COLO checkpoint restore loop */
> +
> + ret = colo_ctl_put(mis->to_src_file, COLO_COMMAND_CHECKPOINT_READY, 0);
> + if (ret < 0) {
> + goto out;
> + }
> +
> + while (mis->state == MIGRATION_STATUS_COLO) {
> + int request = 0;
> + int ret = colo_wait_handle_cmd(mis->from_src_file, &request);
> +
> + if (ret < 0) {
> + break;
> + } else {
> + if (!request) {
> + continue;
> + }
> + }
> +
> + ret = colo_ctl_put(mis->to_src_file, COLO_COMMAND_CHECKPOINT_REPLY, 0);
> + if (ret < 0) {
> + goto out;
> + }
> +
> + ret = colo_ctl_get(mis->from_src_file, COLO_COMMAND_VMSTATE_SEND);
> + if (ret < 0) {
> + goto out;
> + }
> +
> + /* TODO: read migration data into colo buffer */
> +
> + ret = colo_ctl_put(mis->to_src_file, COLO_COMMAND_VMSTATE_RECEIVED, 0);
> + if (ret < 0) {
> + goto out;
> + }
> +
> + /* TODO: load vm state */
> +
> + ret = colo_ctl_put(mis->to_src_file, COLO_COMMAND_VMSTATE_LOADED, 0);
> + if (ret < 0) {
> + goto out;
> + }
> + }
>
> out:
> if (ret < 0) {
> diff --git a/qapi-schema.json b/qapi-schema.json
> index 22251ec..5c4fe6d 100644
> --- a/qapi-schema.json
> +++ b/qapi-schema.json
> @@ -702,6 +702,33 @@
> '*tls-port': 'int', '*cert-subject': 'str' } }
>
> ##
> +# @COLOCommand
> +#
> +# The colo command
> +#
> +# @invalid: unknown command
> +#
> +# @checkpoint-ready: SVM is ready for checkpointing
> +#
> +# @checkpoint-request: PVM tells SVM to prepare for new checkpointing
> +#
> +# @checkpoint-reply: SVM gets PVM's checkpoint request
> +#
> +# @vmstate-send: VM's state will be sent by PVM.
> +#
> +# @vmstate-size: The total size of VMstate.
> +#
> +# @vmstate-received: VM's state has been received by SVM
> +#
> +# @vmstate-loaded: VM's state has been loaded by SVM
> +#
> +# Since: 2.5
> +##
> +{ 'enum': 'COLOCommand',
> + 'data': [ 'invalid', 'checkpoint-ready', 'checkpoint-request',
> + 'checkpoint-reply', 'vmstate-send', 'vmstate-size',
> + 'vmstate-received', 'vmstate-loaded' ] }
> +
> # @MouseInfo:
> #
> # Information about a mouse device.
> diff --git a/trace-events b/trace-events
> index 9cd6391..ee4679c 100644
> --- a/trace-events
> +++ b/trace-events
> @@ -1499,6 +1499,8 @@ rdma_start_outgoing_migration_after_rdma_source_init(void) ""
>
> # migration/colo.c
> colo_vm_state_change(const char *old, const char *new) "Change '%s' => '%s'"
> +colo_ctl_put(const char *msg, uint64_t value) "Send '%s' cmd, value: %" PRIu64""
> +colo_ctl_get(const char *msg) "Receive '%s' cmd"
>
> # kvm-all.c
> kvm_ioctl(int type, void *arg) "type 0x%x, arg %p"
> --
> 1.8.3.1
>
>
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK
next prev parent reply other threads:[~2015-11-06 18:26 UTC|newest]
Thread overview: 100+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-11-03 11:56 [Qemu-devel] [PATCH COLO-Frame v10 00/38] COarse-grain LOck-stepping(COLO) Virtual Machines for Non-stop Service (FT) zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 01/38] configure: Add parameter for configure to enable/disable COLO support zhanghailiang
2015-11-05 14:52 ` Eric Blake
2015-11-06 7:36 ` zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 02/38] migration: Introduce capability 'x-colo' to migration zhanghailiang
2015-11-13 16:01 ` Eric Blake
2015-11-16 8:35 ` zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 03/38] COLO: migrate colo related info to secondary node zhanghailiang
2015-11-06 16:36 ` Dr. David Alan Gilbert
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 04/38] migration: Add state records for migration incoming zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 05/38] migration: Integrate COLO checkpoint process into migration zhanghailiang
2015-11-06 16:48 ` Dr. David Alan Gilbert
2015-11-13 16:42 ` Eric Blake
2015-11-16 13:00 ` zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 06/38] migration: Integrate COLO checkpoint process into loadvm zhanghailiang
2015-11-06 17:29 ` Dr. David Alan Gilbert
2015-11-09 6:09 ` zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 07/38] migration: Rename the'file' member of MigrationState and MigrationIncomingState zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 08/38] COLO/migration: establish a new communication path from destination to source zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 09/38] COLO: Implement colo checkpoint protocol zhanghailiang
2015-11-06 18:26 ` Dr. David Alan Gilbert [this message]
2015-11-09 6:51 ` zhanghailiang
2015-11-09 7:33 ` zhanghailiang
2015-11-13 16:46 ` Eric Blake
2015-11-17 7:04 ` zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 10/38] COLO: Add a new RunState RUN_STATE_COLO zhanghailiang
2015-11-06 18:28 ` Dr. David Alan Gilbert
2015-11-13 16:47 ` Eric Blake
2015-11-17 7:15 ` zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 11/38] QEMUSizedBuffer: Introduce two help functions for qsb zhanghailiang
2015-11-06 18:30 ` Dr. David Alan Gilbert
2015-11-09 8:14 ` zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 12/38] COLO: Save PVM state to secondary side when do checkpoint zhanghailiang
2015-11-06 18:59 ` Dr. David Alan Gilbert
2015-11-09 9:17 ` zhanghailiang
2015-11-13 18:53 ` Dr. David Alan Gilbert
2015-11-17 10:20 ` zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 13/38] COLO: Load PVM's dirty pages into SVM's RAM cache temporarily zhanghailiang
2015-11-13 15:39 ` Dr. David Alan Gilbert
2015-11-16 7:57 ` zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 14/38] COLO: Load VMState into qsb before restore it zhanghailiang
2015-11-13 16:02 ` Dr. David Alan Gilbert
2015-11-16 8:46 ` zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 15/38] ram/COLO: Record pages received from PVM by re-using migration dirty bitmap zhanghailiang
2015-11-13 16:19 ` Dr. David Alan Gilbert
2015-11-16 9:07 ` zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 16/38] COLO: Flush PVM's cached RAM into SVM's memory zhanghailiang
2015-11-13 16:38 ` Dr. David Alan Gilbert
2015-11-16 12:46 ` zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 17/38] COLO: synchronize PVM's state to SVM periodically zhanghailiang
2015-11-13 18:34 ` Dr. David Alan Gilbert
2015-11-17 9:11 ` zhanghailiang
2015-11-17 10:08 ` Dr. David Alan Gilbert
2015-11-17 10:29 ` zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 18/38] COLO failover: Introduce a new command to trigger a failover zhanghailiang
2015-11-13 16:59 ` Eric Blake
2015-11-17 8:03 ` zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 19/38] COLO failover: Introduce state to record failover process zhanghailiang
2015-11-20 15:51 ` Dr. David Alan Gilbert
2015-11-23 5:56 ` zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 20/38] COLO: Implement failover work for Primary VM zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 21/38] COLO: Implement failover work for Secondary VM zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 22/38] COLO: implement default failover treatment zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 23/38] qmp event: Add event notification for COLO error zhanghailiang
2015-11-20 21:50 ` Eric Blake
2015-11-23 6:01 ` zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 24/38] COLO failover: Shutdown related socket fd when do failover zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 25/38] COLO failover: Don't do failover during loading VM's state zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 26/38] COLO: Control the checkpoint delay time by migrate-set-parameters command zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 27/38] COLO: Process shutdown command for VM in COLO state zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 28/38] COLO: Update the global runstate after going into colo state zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 29/38] savevm: Split load vm state function qemu_loadvm_state zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 30/38] COLO: Separate the process of saving/loading ram and device state zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 31/38] COLO: Split qemu_savevm_state_begin out of checkpoint process zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 32/38] netfilter: Add a public API to release all the buffered packets zhanghailiang
2015-11-03 12:39 ` Yang Hongyang
2015-11-03 13:19 ` zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 33/38] netfilter: Introduce an API to delete the timer of all buffer-filters zhanghailiang
2015-11-03 12:41 ` Yang Hongyang
2015-11-03 13:07 ` zhanghailiang
2015-11-04 2:51 ` Jason Wang
2015-11-04 3:08 ` zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 34/38] filter-buffer: Accept zero interval zhanghailiang
2015-11-03 12:43 ` Yang Hongyang
2015-11-04 2:52 ` Jason Wang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 35/38] netfilter: Introduce a API to automatically add filter-buffer for each netdev zhanghailiang
2015-11-03 12:57 ` Yang Hongyang
2015-11-03 13:16 ` zhanghailiang
2015-11-04 2:56 ` Jason Wang
2015-11-04 3:07 ` zhanghailiang
2015-11-05 7:43 ` zhanghailiang
2015-11-05 8:52 ` Wen Congyang
2015-11-05 9:21 ` Jason Wang
2015-11-05 9:33 ` Wen Congyang
2015-11-05 9:19 ` Jason Wang
2015-11-05 10:58 ` zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 36/38] netfilter: Introduce an API to delete all the automatically added netfilters zhanghailiang
2015-11-03 12:58 ` Yang Hongyang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 37/38] colo: Use the netfilter to buffer and release packets zhanghailiang
2015-11-03 11:56 ` [Qemu-devel] [PATCH COLO-Frame v10 38/38] COLO: Add block replication into colo process zhanghailiang
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20151106182632.GH4267@work-vm \
--to=dgilbert@redhat.com \
--cc=amit.shah@redhat.com \
--cc=arei.gonglei@huawei.com \
--cc=eddie.dong@intel.com \
--cc=lizhijian@cn.fujitsu.com \
--cc=peter.huangpeng@huawei.com \
--cc=qemu-devel@nongnu.org \
--cc=quintela@redhat.com \
--cc=stefanha@redhat.com \
--cc=yunhong.jiang@intel.com \
--cc=zhang.zhanghailiang@huawei.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).