Skip to contentSkip to conversation

nvme-fabrics: reject I/O to offline device

PATCH Hannes Reinecke 1 participant 1 message Sep 30, 2020 1 message not in the archive NVMe

lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]

Hannes Reinecke

author Reviewed-by
#1 lore raw
Quoted 54 lines · Victor Gladkov
Quoted 11 lines · Hannes Reinecke

Review

Quoted 7 lines
   		opts->max_reconnects = DIV_ROUND_UP(ctrl_loss_tmo,
   						opts->reconnect_delay);
(quoted diff) +5 -0
+		if (ctrl_loss_tmo < opts->fast_io_fail_tmo)
+			pr_warn("failfast tmo (%d) larger than controller "
+				"loss tmo (%d)\n",
+				opts->fast_io_fail_tmo, ctrl_loss_tmo);
+	}

If we already check for that condition, shouldn't we disable fast_io_fail_tmo in that situation to clarify things?

OK for me. I don't mind

Quoted 25 lines
Quoted 25 lines
   	if (!opts->host) {
   		kref_get(&nvmf_default_host->ref);
(quoted diff) +1 -1
@@ -985,7 +1004,7 @@ void nvmf_free_options(struct nvmf_ctrl_options
 *opts)
   #define NVMF_ALLOWED_OPTS	(NVMF_OPT_QUEUE_SIZE |
 NVMF_OPT_NR_IO_QUEUES | \
   				 NVMF_OPT_KATO | NVMF_OPT_HOSTNQN | \
   				 NVMF_OPT_HOST_ID | NVMF_OPT_DUP_CONNECT
 |\
-				 NVMF_OPT_DISABLE_SQFLOW)
+				 NVMF_OPT_DISABLE_SQFLOW |

NVMF_OPT_FAIL_FAST_TMO)

static struct nvme_ctrl *
nvmf_create_ctrl(struct device *dev, const char *buf) diff --git
a/drivers/nvme/host/fabrics.h b/drivers/nvme/host/fabrics.h index

(quoted diff) +0 -0

b/drivers/nvme/host/multipath.c index 54603bd..d8b7f45 100644

drivers/nvme/host/multipath.c +0 -0
@@ -278,9 +278,12 @@ static bool nvme_available_path(struct
 nvme_ns_head *head)
 
Quoted 12 lines

drivers/nvme/host/multipath.c static bool nvme_available_path(struct · line 285

Quoted 6 lines
drivers/nvme/host/multipath.c +4 -0
   	list_for_each_entry_rcu(ns, &head->list, siblings) {
   		switch (ns->ctrl->state) {
+		case NVME_CTRL_CONNECTING:
+			if (test_bit(NVME_CTRL_FAILFAST_EXPIRED,
+				     &ns->ctrl->flags))
+				break;

No. We shouldn't select this path, but that doesn't mean that all other paths in this list can't be selected, either; they might be coming from different controllers.
Please use 'continue' here.

Quoted 2 lines

The 'break' doesn't interrupt the 'for_each' loop, but only go out from 'switch'

Ah. But still; that is not quite correct as we'll need to intercept things at nvme_ns_head_submit_bio() to make the correct decision there.

I've attached the modified version I'm working with; please check if you're okay with it.

Cheers,

Hannes

Signature · 4 lines
Dr. Hannes Reinecke		           Kernel Storage Architect
[email protected]			                  +49 911 74053 688
SUSE Software Solutions Germany GmbH, Maxfeldstr. 5, 90409 Nürnberg
HRB 36809 (AG Nürnberg), GF: Felix Imendörffer

From fc2e2c5ceff5d0807623ccd83d735965ba5b7ec0 Mon Sep 17 00:00:00 2001
From: Victor Gladkov <[email protected]>
Date: Tue, 29 Sep 2020 15:27:38 +0000
Subject: [PATCH] nvme-fabrics: reject I/O to offline device

Commands get stuck while Host NVMe-oF controller is in reconnect state. NVMe controller enters into reconnect state when it loses connection with the target. It tries to reconnect every 10 seconds
(default) until successful reconnection or until reconnect time-out is reached. The default reconnect time out is 10 minutes.

Applications are expecting commands to complete with success or error within a certain timeout (30 seconds by default). The NVMe host is enforcing that timeout while it is connected, never the less, during reconnection, the timeout is not enforced and commands may get stuck for a long period or even forever.

To fix this long delay due to the default timeout we introduce new session parameter "fast_io_fail_tmo". The timeout is measured in seconds from the controller reconnect, any command beyond that timeout is rejected. The new parameter value may be passed during 'connect'.
The default value of -1 means no timeout (similar to current behavior).

We add a new controller NVME_CTRL_FAILFAST_EXPIRED and respective delayed work that updates the NVME_CTRL_FAILFAST_EXPIRED flag.

When the controller is entering the CONNECTING state, we schedule the delayed_work based on failfast timeout value. If the transition is out of CONNECTING, terminate delayed work item and ensure failfast_expired is false. If delayed work item expires then set "NVME_CTRL_FAILFAST_EXPIRED" flag to true.

We also update nvmf_fail_nonready_command() and
nvme_available_path() functions with check the
"NVME_CTRL_FAILFAST_EXPIRED" controller flag.

For multipath (function nvme_available_path()):
The path will not be considered available if "NVME_CTRL_FAILFAST_EXPIRED" controller flag is set and the controller in NVME_CTRL_CONNECTING state. This prevents commands from getting stuck
when available paths have tried to reconnect for too long.

Signed-off-by Victor Gladkov <victor.gladkov at kioxia.com>
Signed-off-by Chaitanya Kulkarni <chaitanya.kulkarni at wdc.com>
Reviewed-by Hannes Reinecke <hare at suse.de>
---
 drivers/nvme/host/core.c      | 42 ++++++++++++++++++++++++++++++++++++++++--
 drivers/nvme/host/fabrics.c   | 24 ++++++++++++++++++++++--
 drivers/nvme/host/fabrics.h   |  5 +++++
 drivers/nvme/host/multipath.c |  4 ++++
 drivers/nvme/host/nvme.h      |  3 +++
 5 files changed, 74 insertions(+), 4 deletions(-)
drivers/nvme/host/core.c +40 -2
index 385b10317873..9610d81f2070 100644
@@ -148,6 +148,37 @@ int nvme_try_sched_reset(struct nvme_ctrl *ctrl)
 }
 EXPORT_SYMBOL_GPL(nvme_try_sched_reset);
 
+static void nvme_failfast_work(struct work_struct *work) {
+	struct nvme_ctrl *ctrl = container_of(to_delayed_work(work),
+			struct nvme_ctrl, failfast_work);
+
+	if (ctrl->state != NVME_CTRL_CONNECTING)
+		return;
+
+
+	set_bit(NVME_CTRL_FAILFAST_EXPIRED, &ctrl->flags);
+	dev_info(ctrl->device, "failfast expired\n");
+	nvme_kick_requeue_lists(ctrl);
+}
+
+static inline void nvme_start_failfast_work(struct nvme_ctrl *ctrl)
+{
+	if (!ctrl->opts || ctrl->opts->fast_io_fail_tmo == -1)
+		return;
+
+	schedule_delayed_work(&ctrl->failfast_work,
+			      ctrl->opts->fast_io_fail_tmo * HZ);
+}
+
+static inline void nvme_stop_failfast_work(struct nvme_ctrl *ctrl)
+{
+	if (!ctrl->opts)
+		return;
+
+	cancel_delayed_work_sync(&ctrl->failfast_work);
+	clear_bit(NVME_CTRL_FAILFAST_EXPIRED, &ctrl->flags);
+}
+
 int nvme_reset_ctrl(struct nvme_ctrl *ctrl)
 {
 	if (!nvme_change_ctrl_state(ctrl, NVME_CTRL_RESETTING))
@@ -360,9 +391,11 @@ bool nvme_change_ctrl_state(struct nvme_ctrl *ctrl,
 	switch (new_state) {
 	case NVME_CTRL_LIVE:
 		switch (old_state) {
+		case NVME_CTRL_CONNECTING:
+			nvme_stop_failfast_work(ctrl);
+			fallthrough;
 		case NVME_CTRL_NEW:
 		case NVME_CTRL_RESETTING:
-		case NVME_CTRL_CONNECTING:
 			changed = true;
 			fallthrough;
 		default:
@@ -381,8 +414,10 @@ bool nvme_change_ctrl_state(struct nvme_ctrl *ctrl,
 		break;
 	case NVME_CTRL_CONNECTING:
 		switch (old_state) {
-		case NVME_CTRL_NEW:
 		case NVME_CTRL_RESETTING:
+			nvme_start_failfast_work(ctrl);
+			fallthrough;
+		case NVME_CTRL_NEW:
 			changed = true;
 			fallthrough;
 		default:
@@ -4343,6 +4378,7 @@ void nvme_stop_ctrl(struct nvme_ctrl *ctrl)
 {
 	nvme_mpath_stop(ctrl);
 	nvme_stop_keep_alive(ctrl);
+	nvme_stop_failfast_work(ctrl);
 	flush_work(&ctrl->async_event_work);
 	cancel_work_sync(&ctrl->fw_act_work);
 }
@@ -4408,6 +4444,7 @@ int nvme_init_ctrl(struct nvme_ctrl *ctrl, struct device *dev,
 	int ret;
 
 	ctrl->state = NVME_CTRL_NEW;
+	clear_bit(NVME_CTRL_FAILFAST_EXPIRED, &ctrl->flags);
 	spin_lock_init(&ctrl->lock);
 	mutex_init(&ctrl->scan_lock);
 	INIT_LIST_HEAD(&ctrl->namespaces);
@@ -4424,6 +4461,7 @@ int nvme_init_ctrl(struct nvme_ctrl *ctrl, struct device *dev,
 	init_waitqueue_head(&ctrl->state_wq);
 
 	INIT_DELAYED_WORK(&ctrl->ka_work, nvme_keep_alive_work);
+	INIT_DELAYED_WORK(&ctrl->failfast_work, nvme_failfast_work);
 	memset(&ctrl->ka_cmd, 0, sizeof(ctrl->ka_cmd));
 	ctrl->ka_cmd.common.opcode = nvme_admin_keep_alive;
 
drivers/nvme/host/fabrics.c +22 -2
index 8575724734e0..6404cbf41c96 100644
@@ -549,6 +549,7 @@ blk_status_t nvmf_fail_nonready_command(struct nvme_ctrl *ctrl,
 {
 	if (ctrl->state != NVME_CTRL_DELETING_NOIO &&
 	    ctrl->state != NVME_CTRL_DEAD &&
+	    !test_bit(NVME_CTRL_FAILFAST_EXPIRED, &ctrl->flags) &&
 	    !blk_noretry_request(rq) && !(rq->cmd_flags & REQ_NVME_MPATH))
 		return BLK_STS_RESOURCE;
 
@@ -615,6 +616,7 @@ static const match_table_t opt_tokens = {
 	{ NVMF_OPT_NR_WRITE_QUEUES,	"nr_write_queues=%d"	},
 	{ NVMF_OPT_NR_POLL_QUEUES,	"nr_poll_queues=%d"	},
 	{ NVMF_OPT_TOS,			"tos=%d"		},
+	{ NVMF_OPT_FAIL_FAST_TMO,	"fast_io_fail_tmo=%d"	},
 	{ NVMF_OPT_ERR,			NULL			}
 };
 
@@ -634,6 +636,7 @@ static int nvmf_parse_options(struct nvmf_ctrl_options *opts,
 	opts->reconnect_delay = NVMF_DEF_RECONNECT_DELAY;
 	opts->kato = NVME_DEFAULT_KATO;
 	opts->duplicate_connect = false;
+	opts->fast_io_fail_tmo = NVMF_DEF_FAIL_FAST_TMO;
 	opts->hdr_digest = false;
 	opts->data_digest = false;
 	opts->tos = -1; /* < 0 == use transport default */
@@ -754,6 +757,17 @@ static int nvmf_parse_options(struct nvmf_ctrl_options *opts,
 				pr_warn("ctrl_loss_tmo < 0 will reconnect forever\n");
 			ctrl_loss_tmo = token;
 			break;
+		case NVMF_OPT_FAIL_FAST_TMO:
+			if (match_int(args, &token)) {
+				ret = -EINVAL;
+				goto out;
+			}
+
+			if (token >= 0)
+				pr_warn("I/O will fail on reconnect "
+					"controller after %d sec\n", token);
+			opts->fast_io_fail_tmo = token;
+			break;
 		case NVMF_OPT_HOSTNQN:
 			if (opts->host) {
 				pr_err("hostnqn already user-assigned: %s\n",
@@ -886,9 +900,14 @@ static int nvmf_parse_options(struct nvmf_ctrl_options *opts,
 	}
 	if (ctrl_loss_tmo < 0)
 		opts->max_reconnects = -1;
-	else
+	else {
 		opts->max_reconnects = DIV_ROUND_UP(ctrl_loss_tmo,
 						opts->reconnect_delay);
+		if (ctrl_loss_tmo < opts->fast_io_fail_tmo)
+			pr_warn("failfast tmo (%d) larger than controller "
+				"dev loss tmo (%d)\n",
+				opts->fast_io_fail_tmo, ctrl_loss_tmo);
+	}
 
 	if (!opts->host) {
 		kref_get(&nvmf_default_host->ref);
@@ -988,7 +1007,8 @@ EXPORT_SYMBOL_GPL(nvmf_free_options);
 #define NVMF_ALLOWED_OPTS	(NVMF_OPT_QUEUE_SIZE | NVMF_OPT_NR_IO_QUEUES | \
 				 NVMF_OPT_KATO | NVMF_OPT_HOSTNQN | \
 				 NVMF_OPT_HOST_ID | NVMF_OPT_DUP_CONNECT |\
-				 NVMF_OPT_DISABLE_SQFLOW)
+				 NVMF_OPT_DISABLE_SQFLOW | \
+				 NVMF_OPT_FAIL_FAST_TMO)
 
 static struct nvme_ctrl *
 nvmf_create_ctrl(struct device *dev, const char *buf)
drivers/nvme/host/fabrics.h +5 -0
index a9c1e3b4585e..2ce0451c00ff 100644
@@ -15,6 +15,8 @@
 #define NVMF_DEF_RECONNECT_DELAY	10
 /* default to 600 seconds of reconnect attempts before giving up */
 #define NVMF_DEF_CTRL_LOSS_TMO		600
+/* default is -1: the fail fast mechanism is disabled */
+#define NVMF_DEF_FAIL_FAST_TMO		-1
 
 /*
  * Define a host as seen by the target.  We allocate one at boot, but also
@@ -56,6 +58,7 @@ enum {
 	NVMF_OPT_NR_WRITE_QUEUES = 1 << 17,
 	NVMF_OPT_NR_POLL_QUEUES = 1 << 18,
 	NVMF_OPT_TOS		= 1 << 19,
+	NVMF_OPT_FAIL_FAST_TMO	= 1 << 20,
 };
 
 /**
@@ -89,6 +92,7 @@ enum {
  * @nr_write_queues: number of queues for write I/O
  * @nr_poll_queues: number of queues for polling I/O
  * @tos: type of service
+ * @fast_io_fail_tmo: Fast I/O fail timeout in seconds
  */
 struct nvmf_ctrl_options {
 	unsigned		mask;
@@ -111,6 +115,7 @@ struct nvmf_ctrl_options {
 	unsigned int		nr_write_queues;
 	unsigned int		nr_poll_queues;
 	int			tos;
+	int			fast_io_fail_tmo;
 };
 
 /*
drivers/nvme/host/multipath.c +4 -0
index 74896be40c17..0e08ca6e0264 100644
@@ -157,6 +157,8 @@ static bool nvme_path_is_disabled(struct nvme_ns *ns)
 	if (test_bit(NVME_NS_ANA_PENDING, &ns->flags) ||
 	    test_bit(NVME_NS_REMOVING, &ns->flags))
 		return true;
+	if (test_bit(NVME_CTRL_FAILFAST_EXPIRED, &ns->ctrl->flags))
+		return true;
 	return false;
 }
 
@@ -279,6 +281,8 @@ static bool nvme_available_path(struct nvme_ns_head *head)
 	struct nvme_ns *ns;
 
 	list_for_each_entry_rcu(ns, &head->list, siblings) {
+		if (test_bit(NVME_CTRL_FAILFAST_EXPIRED, &ns->ctrl->flags))
+			continue;
 		switch (ns->ctrl->state) {
 		case NVME_CTRL_LIVE:
 		case NVME_CTRL_RESETTING:
drivers/nvme/host/nvme.h +3 -0
index 566776100126..37be7d16ab15 100644
@@ -304,6 +304,7 @@ struct nvme_ctrl {
 	struct work_struct scan_work;
 	struct work_struct async_event_work;
 	struct delayed_work ka_work;
+	struct delayed_work failfast_work;
 	struct nvme_command ka_cmd;
 	struct work_struct fw_act_work;
 	unsigned long events;
@@ -337,6 +338,8 @@ struct nvme_ctrl {
 	u16 icdoff;
 	u16 maxcmd;
 	int nr_reconnects;
+	unsigned long flags;
+#define NVME_CTRL_FAILFAST_EXPIRED	0
 	struct nvmf_ctrl_options *opts;
 
 	struct page *discard_page;
Signature · 1 line
2.16.4

_______________________________________________
Linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme