linux-nfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Jeff Layton <jlayton@redhat.com>
To: bfields@fieldses.org
Cc: linux-nfs@vger.kernel.org
Subject: [PATCH 5/6] nfsd: keep stats on worst hash balancing seen so far
Date: Tue, 19 Mar 2013 09:11:45 -0400	[thread overview]
Message-ID: <1363698706-22036-6-git-send-email-jlayton@redhat.com> (raw)
In-Reply-To: <1363698706-22036-1-git-send-email-jlayton@redhat.com>

The typical case with the DRC is a cache miss, so if we keep track of
the max number of entries that we've ever walked over in a search, then
we should have a reasonable estimate of the longest hash chain that
we've ever seen.

With that, we'll also keep track of the total size of the cache when we
see the longest chain. In the case of a tie, we prefer to track the
smallest total cache size in order to properly gauge the worst-case
ratio of max vs. avg chain length.

Signed-off-by: Jeff Layton <jlayton@redhat.com>
---
 fs/nfsd/nfscache.c | 31 +++++++++++++++++++++++++++----
 1 file changed, 27 insertions(+), 4 deletions(-)

diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c
index fd81ca7..17cb0d6 100644
--- a/fs/nfsd/nfscache.c
+++ b/fs/nfsd/nfscache.c
@@ -41,6 +41,12 @@ static unsigned int		payload_misses;
 /* amount of memory (in bytes) currently consumed by the DRC */
 static unsigned int		drc_mem_usage;
 
+/* longest hash chain seen */
+static unsigned int		longest_chain;
+
+/* size of cache when we saw the longest hash chain */
+static unsigned int		longest_chain_cachesize;
+
 /*
  * Calculate the hash index from an XID.
  */
@@ -319,15 +325,30 @@ nfsd_cache_match(struct svc_rqst *rqstp, __wsum csum, struct svc_cacherep *rp)
 static struct svc_cacherep *
 nfsd_cache_search(struct svc_rqst *rqstp, __wsum csum)
 {
-	struct svc_cacherep	*rp;
+	struct svc_cacherep	*rp, *ret = NULL;
 	struct hlist_head 	*rh;
+	unsigned int		entries = 0;
 
 	rh = &cache_hash[request_hash(rqstp->rq_xid)];
 	hlist_for_each_entry(rp, rh, c_hash) {
-		if (nfsd_cache_match(rqstp, csum, rp))
-			return rp;
+		++entries;
+		if (nfsd_cache_match(rqstp, csum, rp)) {
+			ret = rp;
+			break;
+		}
+	}
+
+	/* tally hash chain length stats */
+	if (entries > longest_chain) {
+		longest_chain = entries;
+		longest_chain_cachesize = num_drc_entries;
+	} else if (entries == longest_chain) {
+		/* prefer to keep the smallest cachesize possible here */
+		longest_chain_cachesize = min(longest_chain_cachesize,
+						num_drc_entries);
 	}
-	return NULL;
+
+	return ret;
 }
 
 /*
@@ -573,6 +594,8 @@ static int nfsd_reply_cache_stats_show(struct seq_file *m, void *v)
 	seq_printf(m, "cache misses:          %u\n", nfsdstats.rcmisses);
 	seq_printf(m, "not cached:            %u\n", nfsdstats.rcnocache);
 	seq_printf(m, "payload misses:        %u\n", payload_misses);
+	seq_printf(m, "longest chain len:     %u\n", longest_chain);
+	seq_printf(m, "cachesize at longest:  %u\n", longest_chain_cachesize);
 	spin_unlock(&cache_lock);
 	return 0;
 }
-- 
1.7.11.7


  parent reply	other threads:[~2013-03-19 13:11 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-03-19 13:11 [PATCH 0/6] nfsd: duplicate reply cache improvements for 3.10 Jeff Layton
2013-03-19 13:11 ` [PATCH 1/6] nfsd: eliminate one of the DRC cache searches Jeff Layton
2013-03-19 13:11 ` [PATCH 2/6] nfsd: break out comparator into separate function Jeff Layton
2013-03-19 13:11 ` [PATCH 3/6] nfsd: track memory utilization by the DRC Jeff Layton
2013-03-19 13:11 ` [PATCH 4/6] nfsd: add new reply_cache_stats file in nfsdfs Jeff Layton
2013-03-19 13:11 ` Jeff Layton [this message]
2013-03-19 13:11 ` [PATCH 6/6] nfsd: scale up the number of DRC hash buckets with cache size Jeff Layton

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=1363698706-22036-6-git-send-email-jlayton@redhat.com \
    --to=jlayton@redhat.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).