linux-nfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Chuck Lever <chuck.lever@oracle.com>
To: bfields@fieldses.org
Cc: linux-nfs@vger.kernel.org
Subject: [PATCH v2 09/18] sunrpc: Report per-RPC execution stats
Date: Tue, 27 Mar 2018 10:51:39 -0400	[thread overview]
Message-ID: <20180327145139.7710.14287.stgit@oracle-ib-101.nfsv4bat.org> (raw)
In-Reply-To: <20180327144420.7710.82288.stgit@oracle-ib-101.nfsv4bat.org>

Introduce a mechanism to report the server-side execution latency of
each RPC. The goal is to enable user space to filter the trace
record for latency outliers, build histograms, etc.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
---
 include/linux/sunrpc/svc.h    |    1 +
 include/trace/events/sunrpc.h |   21 +++++++++++++++++++++
 net/sunrpc/svc_xprt.c         |    3 ++-
 3 files changed, 24 insertions(+), 1 deletion(-)

diff --git a/include/linux/sunrpc/svc.h b/include/linux/sunrpc/svc.h
index 786ae22..3bd7504 100644
--- a/include/linux/sunrpc/svc.h
+++ b/include/linux/sunrpc/svc.h
@@ -283,6 +283,7 @@ struct svc_rqst {
 	int			rq_reserved;	/* space on socket outq
 						 * reserved for this request
 						 */
+	ktime_t			rq_stime;	/* start time */
 
 	struct cache_req	rq_chandle;	/* handle passed to caches for 
 						 * request delaying 
diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h
index 5a8157c..1b383f7 100644
--- a/include/trace/events/sunrpc.h
+++ b/include/trace/events/sunrpc.h
@@ -705,6 +705,27 @@
 		__entry->len, show_svc_xprt_flags(__entry->flags))
 );
 
+TRACE_EVENT(svc_stats_latency,
+	TP_PROTO(const struct svc_rqst *rqst),
+
+	TP_ARGS(rqst),
+
+	TP_STRUCT__entry(
+		__field(u32, xid)
+		__field(unsigned long, execute)
+		__string(addr, rqst->rq_xprt->xpt_remotebuf)
+	),
+
+	TP_fast_assign(
+		__entry->xid = be32_to_cpu(rqst->rq_xid);
+		__entry->execute = ktime_to_us(ktime_sub(ktime_get(),
+							 rqst->rq_stime));
+		__assign_str(addr, rqst->rq_xprt->xpt_remotebuf);
+	),
+
+	TP_printk("addr=%s xid=0x%08x execute-us=%lu",
+		__get_str(addr), __entry->xid, __entry->execute)
+);
 
 DECLARE_EVENT_CLASS(svc_deferred_event,
 	TP_PROTO(struct svc_deferred_req *dr),
diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c
index f745754..a7425da 100644
--- a/net/sunrpc/svc_xprt.c
+++ b/net/sunrpc/svc_xprt.c
@@ -782,7 +782,7 @@ static int svc_handle_xprt(struct svc_rqst *rqstp, struct svc_xprt *xprt)
 			len = svc_deferred_recv(rqstp);
 		else
 			len = xprt->xpt_ops->xpo_recvfrom(rqstp);
-		dprintk("svc: got len=%d\n", len);
+		rqstp->rq_stime = ktime_get();
 		rqstp->rq_reserved = serv->sv_max_mesg;
 		atomic_add(rqstp->rq_reserved, &xprt->xpt_reserved);
 	}
@@ -888,6 +888,7 @@ int svc_send(struct svc_rqst *rqstp)
 
 	/* Grab mutex to serialize outgoing data. */
 	mutex_lock(&xprt->xpt_mutex);
+	trace_svc_stats_latency(rqstp);
 	if (test_bit(XPT_DEAD, &xprt->xpt_flags)
 			|| test_bit(XPT_CLOSE, &xprt->xpt_flags))
 		len = -ENOTCONN;


  parent reply	other threads:[~2018-03-27 14:52 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-03-27 14:49 [PATCH v2 00/18] NFS/RDMA server for v4.17 Chuck Lever
2018-03-27 14:49 ` [PATCH v2 01/18] sunrpc: Remove unneeded pointer dereference Chuck Lever
2018-03-27 14:49 ` [PATCH v2 02/18] svc: Simplify ->xpo_secure_port Chuck Lever
2018-03-27 14:49 ` [PATCH v2 03/18] sunrpc: Update show_svc_xprt_flags() to include recently added flags Chuck Lever
2018-03-27 14:50 ` [PATCH v2 04/18] sunrpc: Move trace_svc_xprt_dequeue() Chuck Lever
2018-03-27 14:50 ` [PATCH v2 05/18] sunrpc: Simplify do_enqueue tracing Chuck Lever
2018-03-27 14:50 ` [PATCH v2 06/18] sunrpc: Simplify trace_svc_recv Chuck Lever
2018-03-27 14:51 ` [PATCH v2 07/18] sunrpc: Save remote presentation address in svc_xprt for trace events Chuck Lever
2018-03-27 14:51 ` [PATCH v2 08/18] sunrpc: Re-purpose trace_svc_process Chuck Lever
2018-03-27 14:51 ` Chuck Lever [this message]
2018-03-27 14:52 ` [PATCH v2 10/18] svc: Report xprt dequeue latency Chuck Lever
2018-03-27 14:52 ` [PATCH v2 11/18] nfsd: Fix NFSD trace points Chuck Lever
2018-03-27 14:52 ` [PATCH v2 12/18] nfsd: Record request byte count, not count of vectors Chuck Lever
2018-03-27 14:53 ` [PATCH v2 13/18] nfsd: Add "nfsd_" to trace point names Chuck Lever
2018-03-27 14:53 ` [PATCH v2 14/18] nfsd: Add I/O trace points in the NFSv4 write path Chuck Lever
2018-03-27 14:53 ` [PATCH v2 15/18] nfsd: Add I/O trace points in the NFSv4 read proc Chuck Lever
2018-03-27 16:57   ` Chuck Lever
2018-03-27 20:14     ` Bruce Fields
2018-03-27 21:22       ` Chuck Lever
2018-03-27 21:51         ` Bruce Fields
2018-03-27 14:53 ` [PATCH v2 16/18] nfsd: Trace NFSv4 COMPOUND execution Chuck Lever
2018-03-27 14:54 ` [PATCH v2 17/18] NFSD: Clean up legacy NFS WRITE argument XDR decoders Chuck Lever
2018-03-27 14:54 ` [PATCH v2 18/18] NFSD: Clean up legacy NFS SYMLINK " Chuck Lever

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=20180327145139.7710.14287.stgit@oracle-ib-101.nfsv4bat.org \
    --to=chuck.lever@oracle.com \
    --cc=bfields@fieldses.org \
    --cc=linux-nfs@vger.kernel.org \
    /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).