From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from gandalf.ozlabs.org (mail.ozlabs.org [IPv6:2404:9400:2221:ea00::3]) by passt.top (Postfix) with ESMTPS id 449455A027E for ; Tue, 22 Aug 2023 04:11:44 +0200 (CEST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gibson.dropbear.id.au; s=201602; t=1692670296; bh=o0mKZb2RbLmOK7hWYcoykOkwtuCTPNpGbl1hBrOgCfQ=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=hfilRsCae8WO1IkJzg6/w6vlGcuIWKKTpAoKV1rNzE8Fsg+7YSWP2qIUgfdL/RQik mVjhEDTFvxJlGthMpYOGL+NfyZ8s47eZTv7TtygEr9wpCT8cJrlEeoi/oXr1FcYJxz iyUMq+0BnrdlIiAsybLodXXzdQXYP52mRoK2zsUY= Received: by gandalf.ozlabs.org (Postfix, from userid 1007) id 4RVCWr3M3Sz4x2G; Tue, 22 Aug 2023 12:11:36 +1000 (AEST) From: David Gibson To: Stefano Brivio , passt-dev@passt.top Subject: [PATCH v3 9/9] tcp: Remove broken pressure calculations for tcp_defer_handler() Date: Tue, 22 Aug 2023 12:11:30 +1000 Message-ID: <20230822021130.450542-10-david@gibson.dropbear.id.au> X-Mailer: git-send-email 2.41.0 In-Reply-To: <20230822021130.450542-1-david@gibson.dropbear.id.au> References: <20230822021130.450542-1-david@gibson.dropbear.id.au> MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Message-ID-Hash: DL2WWV225KJYUNR7MZJVZ76LDTYSDBXH X-Message-ID-Hash: DL2WWV225KJYUNR7MZJVZ76LDTYSDBXH X-MailFrom: dgibson@gandalf.ozlabs.org X-Mailman-Rule-Misses: dmarc-mitigation; no-senders; approved; emergency; loop; banned-address; member-moderation; nonmember-moderation; administrivia; implicit-dest; max-recipients; max-size; news-moderation; no-subject; digests; suspicious-header CC: David Gibson X-Mailman-Version: 3.3.8 Precedence: list List-Id: Development discussion and patches for passt Archived-At: Archived-At: List-Archive: List-Archive: List-Help: List-Owner: List-Post: List-Subscribe: List-Unsubscribe: tcp_defer_handler() performs a potentially expensive linear scan of the connection table. So, to mitigate the cost of that we skip if if we're not under at least moderate pressure: either 30% of available connections or 30% (estimated) of available fds used. But, the calculation for this has been broken since it was introduced: we calculate "max_conns" based on c->tcp.conn_count, not TCP_MAX_CONNS, meaning we only exit early if conn_count is less than 30% of itself, i.e. never. If that calculation is "corrected" to be based on TCP_MAX_CONNS, it completely tanks the TCP CRR times for passt - from ~60ms to >1000ms on my laptop. My guess is that this is because in the case of many short lived connections, we're letting the table become much fuller before compacting it. That means that other places which perform a table scan now have to do much, much more. For the time being, simply remove the tests, since they're not doing anything useful. We can reintroduce them more carefully if we see a need for them. Signed-off-by: David Gibson --- tcp.c | 9 --------- 1 file changed, 9 deletions(-) diff --git a/tcp.c b/tcp.c index f396ede..c89e6e4 100644 --- a/tcp.c +++ b/tcp.c @@ -309,9 +309,6 @@ #define TCP_FRAMES \ (c->mode == MODE_PASST ? TCP_FRAMES_MEM : 1) -#define TCP_FILE_PRESSURE 30 /* % of c->nofile */ -#define TCP_CONN_PRESSURE 30 /* % of c->tcp.conn_count */ - #define TCP_HASH_TABLE_LOAD 70 /* % */ #define TCP_HASH_TABLE_SIZE (TCP_MAX_CONNS * 100 / \ TCP_HASH_TABLE_LOAD) @@ -1385,17 +1382,11 @@ static void tcp_l2_data_buf_flush(struct ctx *c) */ void tcp_defer_handler(struct ctx *c) { - int max_conns = c->tcp.conn_count / 100 * TCP_CONN_PRESSURE; - int max_files = c->nofile / 100 * TCP_FILE_PRESSURE; union tcp_conn *conn; tcp_l2_flags_buf_flush(c); tcp_l2_data_buf_flush(c); - if ((c->tcp.conn_count < MIN(max_files, max_conns)) && - (c->tcp.splice_conn_count < MIN(max_files / 6, max_conns))) - return; - for (conn = tc + c->tcp.conn_count - 1; conn >= tc; conn--) { if (conn->c.spliced) { if (conn->splice.flags & CLOSING) -- 2.41.0