Add tuned profile with recommended vm tunings for scoutfs.

Install a scoutfs `tuned` profile as a system-profile for `tuned`
with recommendation hint to enable this.

The goal is to provide a set of basic VM tunings that are reasonably
good starting point for production deployment of scoutfs. The tunings
are chosen to reflect good practices to aim for responsiveness of
the scoutfs deployment.

The values chosen are based on existing tuned profiles, building on
throughput-performance and network-throughput as a base, and tuning
VM values in the same way that latency-performance does. All of them
enable the performance CPU governor. None of this enables powersave
settings.

Different deployments may have different performance characteristics
and require further adjustment, or even a completely different profile.

Signed-off-by: Auke Kok <auke.kok@versity.com>

kmod/src/server.c

@@ -1618,7 +1618,8 @@ static int server_get_log_trees(struct super_block *sb,
 		goto update;
 	}
 
-	ret = alloc_move_empty(sb, &super->data_alloc, &lt.data_freed, 100);
+	ret = alloc_move_empty(sb, &super->data_alloc, &lt.data_freed,
+			       COMMIT_HOLD_ALLOC_BUDGET / 2);
 	if (ret == -EINPROGRESS)
 		ret = 0;
 	if (ret < 0) {
@@ -1913,9 +1914,11 @@ static int reclaim_open_log_tree(struct super_block *sb, u64 rid)
 	       scoutfs_alloc_splice_list(sb, &server->alloc, &server->wri, server->other_freed,
 					 &lt.meta_avail)) ?:
 	      (err_str = "empty data_avail",
-	       alloc_move_empty(sb, &super->data_alloc, &lt.data_avail, 100)) ?:
+	       alloc_move_empty(sb, &super->data_alloc, &lt.data_avail,
+				COMMIT_HOLD_ALLOC_BUDGET / 2)) ?:
 	      (err_str = "empty data_freed",
-	       alloc_move_empty(sb, &super->data_alloc, &lt.data_freed, 100));
+	       alloc_move_empty(sb, &super->data_alloc, &lt.data_freed,
+				COMMIT_HOLD_ALLOC_BUDGET / 2));
 	mutex_unlock(&server->alloc_mutex);
 
 	/* only finalize, allowing merging, once the allocators are fully freed */
@@ -3036,7 +3039,13 @@ static int server_commit_log_merge(struct super_block *sb,
 				  SCOUTFS_LOG_MERGE_STATUS_ZONE, 0, 0,
 				  &stat, sizeof(stat));
 	if (ret < 0) {
-		err_str = "getting merge status item";
+		/*
+		 * During a retransmission, it's possible that the server
+		 * already committed and resolved this log merge. ENOENT
+		 * is expected in that case.
+		 */
+		if (ret != -ENOENT)
+			err_str = "getting merge status item";
 		goto out;
 	}
 

tests/fenced-local-force-unmount.sh

@@ -9,7 +9,7 @@
 echo "$0 running rid '$SCOUTFS_FENCED_REQ_RID' ip '$SCOUTFS_FENCED_REQ_IP' args '$@'"
 
 echo_fail() {
-	echo "$@" >> /dev/stderr
+	echo "$@" >&2
 	exit 1
 }
 
@@ -27,8 +27,7 @@ for fs in /sys/fs/scoutfs/*; do
 	nr="$(quiet_cat $fs/data_device_maj_min)"
 	[ ! -d "$fs" -o "$fs_rid" != "$rid" ] && continue
 
-	mnt=$(findmnt -l -n -t scoutfs -o TARGET -S $nr) || \
-		echo_fail "findmnt -t scoutfs -S $nr failed"
+	mnt=$(findmnt -l -n -t scoutfs -o TARGET -S $nr)
 	[ -z "$mnt" ] && continue
 
 	if ! umount -qf "$mnt"; then

tests/funcs/filter.sh

@@ -170,6 +170,9 @@ t_filter_dmesg()
 	# some ci test guests are unresponsive
 	re="$re|longest quorum heartbeat .* delay"
 
+	# creating block devices may trigger this
+	re="$re|block device autoloading is deprecated and will be removed."
+
 	egrep -v "($re)" | \
 		ignore_harmless_unwind_kasan_stack_oob
 }

tests/funcs/tap.sh

@@ -43,9 +43,14 @@ t_tap_progress()
 	local testname=$1
 	local result=$2
 
+	local stmsg=""
 	local diff=""
 	local dmsg=""
 
+	if [[ -s $T_RESULTS/tmp/${testname}/status.msg ]]; then
+		stmsg="1"
+	fi
+
 	if [[ -s "$T_RESULTS/tmp/${testname}/dmesg.new" ]]; then
 		dmsg="1"
 	fi
@@ -61,6 +66,7 @@ t_tap_progress()
 		echo "# ${testname} ** skipped - permitted **"
 	else
 		echo "not ok ${i} - ${testname}"
+
 		case ${result} in
 		101)
 			echo "# ${testname} ** skipped **"
@@ -70,6 +76,13 @@ t_tap_progress()
 			;;
 		esac
 
+		if [[ -n "${stmsg}" ]]; then
+			echo "#"
+			echo "# status:"
+			echo "#"
+			cat $T_RESULTS/tmp/${testname}/status.msg | sed 's/^/# - /'
+		fi
+
 		if [[ -n "${diff}" ]]; then
 			echo "#"
 			echo "# diff:"

tests/tests/get-referring-entries.sh

@@ -72,7 +72,7 @@ touch $T_D0/dir/file
 mkdir $T_D0/dir/dir
 ln -s $T_D0/dir/file $T_D0/dir/symlink
 mknod $T_D0/dir/char c 1 3 # null
-mknod $T_D0/dir/block b 7 0 # loop0
+mknod $T_D0/dir/block b 42 0 # SAMPLE block dev - nonexistant/demo use only number
 for name in $(ls -UA $T_D0/dir | sort); do
 	ino=$(stat -c '%i' $T_D0/dir/$name)
 	$GRE $ino | filter_types

tests/tests/renameat2-noreplace.sh

@@ -8,19 +8,19 @@ t_require_mounts 2
 echo "=== renameat2 noreplace flag test"
 
 # give each mount their own dir (lock group) to minimize create contention
-mkdir $T_M0/dir0
-mkdir $T_M1/dir1
+mkdir $T_D0/dir0
+mkdir $T_D1/dir1
 
 echo "=== run two asynchronous calls to renameat2 NOREPLACE"
 for i in $(seq 0 100); do
         # prepare inputs in isolation
-        touch "$T_M0/dir0/old0"
-        touch "$T_M1/dir1/old1"
+        touch "$T_D0/dir0/old0"
+        touch "$T_D1/dir1/old1"
 
         # race doing noreplace renames, both can't succeed
-        dumb_renameat2 -n "$T_M0/dir0/old0" "$T_M0/dir0/sharednew" 2> /dev/null &
+        dumb_renameat2 -n "$T_D0/dir0/old0" "$T_D0/dir0/sharednew" 2> /dev/null &
         pid0=$!
-        dumb_renameat2 -n "$T_M1/dir1/old1" "$T_M1/dir0/sharednew" 2> /dev/null &
+        dumb_renameat2 -n "$T_D1/dir1/old1" "$T_D1/dir0/sharednew" 2> /dev/null &
         pid1=$!
 
         wait $pid0
@@ -31,7 +31,7 @@ for i in $(seq 0 100); do
         test "$rc0" == 0 -a "$rc1" == 0 && t_fail "both renames succeeded"
 
         # blow away possible files for either race outcome
-        rm -f "$T_M0/dir0/old0" "$T_M1/dir1/old1" "$T_M0/dir0/sharednew" "$T_M1/dir1/sharednew"
+        rm -f "$T_D0/dir0/old0" "$T_D1/dir1/old1" "$T_D0/dir0/sharednew" "$T_D1/dir1/sharednew"
 done
 
 t_pass

utils/fenced/scoutfs-fenced

@@ -7,7 +7,7 @@ message_output()
 
 error_message()
 {
-	message_output "$@" >> /dev/stderr
+	message_output "$@" >&2
 }
 
 error_exit()

utils/man/scoutfs.8

@@ -402,39 +402,25 @@ before destroying an old empty data device.
 .PD
 
 .TP
-.BI "print {-a|--allocs} {-i|--items ITEMS} {-r|--roots ROOTS} {-S|--skip-likely-huge} META-DEVICE"
+.BI "print {-S|--skip-likely-huge} META-DEVICE"
 .sp
-Prints out some or all of the metadata in the file system.  This makes no effort
+Prints out all of the metadata in the file system.  This makes no effort
 to ensure that the structures are consistent as they're traversed and
 can present structures that seem corrupt as they change as they're
 output.
-.sp
-Structures that are related to the number of mounts and are maintained at a
-relatively reasonable size are always printed. These include per-mount log
-trees, srch files, allocators, and the metadata allocators used by server
-commits. Other btrees and their items can be selected as desired.
 .RS 1.0i
 .PD 0
+.TP
 .sp
-.TP
-.B "-a, --allocs"
-Print the metadata and data allocators. Enabled by default.
-.TP
-.B "-r, --roots ROOTS"
-This option can be used to select which btrees are traversed. It is a comma-separated list containing one or more of the following btree roots: logs, srch, fs. Default is all roots.
-.TP
-.B "-i, --items ITEMS"
-This option can be used to choose which btree items are printed from the
-selected btree roots. It is a comma-separated list containing one or
-more of the following items: inode, xattr, dirent, symlink, backref, extent.
-Default is all items.
-.TP
 .B "-S, --skip-likely-huge"
 Skip printing structures that are likely to be very large.  The
 structures that are skipped tend to be global and whose size tends to be
 related to the size of the volume.   Examples of skipped structures include
 the global fs items, srch files, and metadata and data
-allocators.
+allocators.  Similar structures that are not skipped are related to the
+number of mounts and are maintained at a relatively reasonable size.
+These include per-mount log trees, srch files, allocators, and the
+metadata allocators used by server commits.
 .sp
 Skipping the larger structures limits the print output to a relatively
 constant size rather than being a large multiple of the used metadata

utils/scoutfs-utils.spec.in

@@ -4,6 +4,12 @@
 
 %{!?_release: %global _release 0.%{pkg_date}git%{pkg_git_hash}}
 
+%if 0%{?rhel} && 0%{?rhel} < 10
+%global tuned_profiles_dir %{_prefix}/lib/tuned
+%else
+%global tuned_profiles_dir %{_prefix}/lib/tuned/profiles
+%endif
+
 Name:           scoutfs-utils
 Summary:        scoutfs user space utilities
 Version:        %{pkg_version}
@@ -57,6 +63,8 @@ install -m 644 -D src/format.h $RPM_BUILD_ROOT%{_includedir}/scoutfs/format.h
 install -m 755 -D fenced/scoutfs-fenced $RPM_BUILD_ROOT%{_libexecdir}/scoutfs-fenced/scoutfs-fenced
 install -m 644 -D fenced/scoutfs-fenced.service $RPM_BUILD_ROOT%{_unitdir}/scoutfs-fenced.service
 install -m 644 -D fenced/scoutfs-fenced.conf.example $RPM_BUILD_ROOT%{_sysconfdir}/scoutfs/scoutfs-fenced.conf.example
+install -m 644 -D tuned/tuned.conf $RPM_BUILD_ROOT%{tuned_profiles_dir}/scoutfs/tuned.conf
+install -m 644 -D tuned/40-scoutfs.conf $RPM_BUILD_ROOT%{_prefix}/lib/tuned/recommend.d/40-scoutfs.conf
 
 %files
 %defattr(644,root,root,755)
@@ -66,6 +74,8 @@ install -m 644 -D fenced/scoutfs-fenced.conf.example $RPM_BUILD_ROOT%{_sysconfdi
 %defattr(755,root,root,755)
 %{_sbindir}/scoutfs
 %{_libexecdir}/scoutfs-fenced
+%{tuned_profiles_dir}/scoutfs/tuned.conf
+%{_prefix}/lib/tuned/recommend.d/40-scoutfs.conf
 
 %files -n scoutfs-devel
 %defattr(644,root,root,755)

utils/src/print.c

@@ -29,42 +29,6 @@
 #include "leaf_item_hash.h"
 #include "dev.h"
 
-struct print_args {
-	char *meta_device;
-	bool skip_likely_huge;
-	bool roots_requested;
-	bool items_requested;
-	bool allocs_requested;
-	bool walk_allocs;
-	bool walk_logs_root;
-	bool walk_fs_root;
-	bool walk_srch_root;
-	bool print_inodes;
-	bool print_xattrs;
-	bool print_dirents;
-	bool print_symlinks;
-	bool print_backrefs;
-	bool print_extents;
-};
-
-static struct print_args print_args = {
-	.meta_device	  = NULL,
-	.skip_likely_huge = false,
-	.roots_requested  = false,
-	.items_requested  = false,
-	.allocs_requested = false,
-	.walk_allocs	  = true,
-	.walk_logs_root	  = true,
-	.walk_fs_root	  = true,
-	.walk_srch_root	  = true,
-	.print_inodes	  = true,
-	.print_xattrs	  = true,
-	.print_dirents	  = true,
-	.print_symlinks	  = true,
-	.print_backrefs	  = true,
-	.print_extents	  = true
-};
-
 static void print_block_header(struct scoutfs_block_header *hdr, int size)
 {
 	u32 crc = crc_block(hdr, size);
@@ -231,7 +195,7 @@ static void print_inode_index(struct scoutfs_key *key, void *val, int val_len)
 
 typedef void (*print_func_t)(struct scoutfs_key *key, void *val, int val_len);
 
-static print_func_t find_printer(u8 zone, u8 type, bool *suppress)
+static print_func_t find_printer(u8 zone, u8 type)
 {
 	if (zone == SCOUTFS_INODE_INDEX_ZONE &&
 	    type >= SCOUTFS_INODE_INDEX_META_SEQ_TYPE  &&
@@ -254,34 +218,13 @@ static print_func_t find_printer(u8 zone, u8 type, bool *suppress)
 
 	if (zone == SCOUTFS_FS_ZONE) {
 		switch(type) {
-			case SCOUTFS_INODE_TYPE:
-				if (!print_args.print_inodes)
-					*suppress = true;
-				return print_inode;
-			case SCOUTFS_XATTR_TYPE:
-				if (!print_args.print_xattrs)
-					*suppress = true;
-				return print_xattr;
-			case SCOUTFS_DIRENT_TYPE:
-				if (!print_args.print_dirents)
-					*suppress = true;
-				return print_dirent;
-			case SCOUTFS_READDIR_TYPE:
-				if (!print_args.print_dirents)
-					*suppress = true;
-				return print_dirent;
-			case SCOUTFS_SYMLINK_TYPE:
-				if (!print_args.print_symlinks)
-					*suppress = true;
-				return print_symlink;
-			case SCOUTFS_LINK_BACKREF_TYPE:
-				if (!print_args.print_backrefs)
-					*suppress = true;
-				return print_dirent;
-			case SCOUTFS_DATA_EXTENT_TYPE:
-				if (!print_args.print_extents)
-					*suppress = true;
-				return print_data_extent;
+			case SCOUTFS_INODE_TYPE: return print_inode;
+			case SCOUTFS_XATTR_TYPE: return print_xattr;
+			case SCOUTFS_DIRENT_TYPE: return print_dirent;
+			case SCOUTFS_READDIR_TYPE: return print_dirent;
+			case SCOUTFS_SYMLINK_TYPE: return print_symlink;
+			case SCOUTFS_LINK_BACKREF_TYPE: return print_dirent;
+			case SCOUTFS_DATA_EXTENT_TYPE: return print_data_extent;
 		}
 	}
 
@@ -301,16 +244,12 @@ static int print_fs_item(struct scoutfs_key *key, u64 seq, u8 flags, void *val,
 
 	/* only items in leaf blocks have values */
 	if (val != NULL && !(flags & SCOUTFS_ITEM_FLAG_DELETION)) {
-		bool suppress = false;
-
-		printer = find_printer(key->sk_zone, key->sk_type, &suppress);
-		if (printer) {
-			if (!suppress)
-				printer(key, val, val_len);
-		} else {
+		printer = find_printer(key->sk_zone, key->sk_type);
+		if (printer)
+			printer(key, val, val_len);
+		else
 			printf("      (unknown zone %u type %u)\n",
 			       key->sk_zone, key->sk_type);
-		}
 	}
 
 	return 0;
@@ -1098,7 +1037,12 @@ static void print_super_block(struct scoutfs_super_block *super, u64 blkno)
 	}
 }
 
-static int print_volume(int fd)
+struct print_args {
+	char *meta_device;
+	bool skip_likely_huge;
+};
+
+static int print_volume(int fd, struct print_args *args)
 {
 	struct scoutfs_super_block *super = NULL;
 	struct print_recursion_args pa;
@@ -1148,7 +1092,7 @@ static int print_volume(int fd)
 			ret = err;
 	}
 
-	if (print_args.walk_allocs) {
+	if (!args->skip_likely_huge) {
 		for (i = 0; i < array_size(super->meta_alloc); i++) {
 			snprintf(str, sizeof(str), "meta_alloc[%u]", i);
 			err = print_btree(fd, super, str, &super->meta_alloc[i].root,
@@ -1175,21 +1119,18 @@ static int print_volume(int fd)
 
 	pa.super = super;
 	pa.fd = fd;
-	if (print_args.walk_srch_root) {
+	if (!args->skip_likely_huge) {
 		err = print_btree_leaf_items(fd, super, &super->srch_root.ref,
 					     print_srch_root_files, &pa);
 		if (err && !ret)
 			ret = err;
 	}
+	err = print_btree_leaf_items(fd, super, &super->logs_root.ref,
+				     print_log_trees_roots, &pa);
+	if (err && !ret)
+		ret = err;
 
-	if (print_args.walk_logs_root) {
-		err = print_btree_leaf_items(fd, super, &super->logs_root.ref,
-					     print_log_trees_roots, &pa);
-		if (err && !ret)
-			ret = err;
-	}
-
-	if (print_args.walk_fs_root) {
+	if (!args->skip_likely_huge) {
 		err = print_btree(fd, super, "fs_root", &super->fs_root,
 				  print_fs_item, NULL);
 		if (err && !ret)
@@ -1202,16 +1143,16 @@ out:
 	return ret;
 }
 
-static int do_print(void)
+static int do_print(struct print_args *args)
 {
 	int ret;
 	int fd;
 
-	fd = open(print_args.meta_device, O_RDONLY);
+	fd = open(args->meta_device, O_RDONLY);
 	if (fd < 0) {
 		ret = -errno;
 		fprintf(stderr, "failed to open '%s': %s (%d)\n",
-			print_args.meta_device, strerror(errno), errno);
+			args->meta_device, strerror(errno), errno);
 		return ret;
 	}
 
@@ -1219,169 +1160,30 @@ static int do_print(void)
 	if (ret < 0)
 		goto out;
 
-	ret = print_volume(fd);
+	ret = print_volume(fd, args);
 out:
 	close(fd);
 	return ret;
 };
 
-enum {
-	LOGS_OPT = 0,
-	FS_OPT,
-	SRCH_OPT
-};
-
-static char *const root_tokens[] = {
-	[LOGS_OPT] = "logs",
-	[FS_OPT] =   "fs",
-	[SRCH_OPT] = "srch",
-	NULL
-};
-
-enum {
-	INODE_OPT = 0,
-	XATTR_OPT,
-	DIRENT_OPT,
-	SYMLINK_OPT,
-	BACKREF_OPT,
-	EXTENT_OPT
-};
-
-static char *const item_tokens[] = {
-	[INODE_OPT] =   "inode",
-	[XATTR_OPT] =   "xattr",
-	[DIRENT_OPT] =  "dirent",
-	[SYMLINK_OPT] = "symlink",
-	[BACKREF_OPT] = "backref",
-	[EXTENT_OPT] =  "extent",
-	NULL
-};
-
-static void clear_items(void)
-{
-	print_args.print_inodes = false;
-	print_args.print_xattrs = false;
-	print_args.print_dirents = false;
-	print_args.print_symlinks = false;
-	print_args.print_backrefs = false;
-	print_args.print_extents = false;
-}
-
-static void clear_roots(void)
-{
-	print_args.walk_logs_root = false;
-	print_args.walk_fs_root = false;
-	print_args.walk_srch_root = false;
-}
-
 static int parse_opt(int key, char *arg, struct argp_state *state)
 {
 	struct print_args *args = state->input;
-	char *subopts;
-	char *value;
-	bool parse_err = false;
 
 	switch (key) {
 	case 'S':
 		args->skip_likely_huge = true;
 		break;
-
-	case 'a':
-		args->allocs_requested = true;
-		args->walk_allocs = true;
-		break;
-
-	case 'i':
-		/* Specific items being requested- clear them all to start */
-		if (!args->items_requested) {
-			clear_items();
-			if (!args->allocs_requested)
-				args->walk_allocs = false;
-			args->items_requested = true;
-		}
-
-		subopts = arg;
-		while (*subopts != '\0' && !parse_err) {
-			switch (getsubopt(&subopts, item_tokens, &value)) {
-			case INODE_OPT:
-				args->print_inodes = true;
-				break;
-			case XATTR_OPT:
-				args->print_xattrs = true;
-				break;
-			case DIRENT_OPT:
-				args->print_dirents = true;
-				break;
-			case SYMLINK_OPT:
-				args->print_symlinks = true;
-				break;
-			case BACKREF_OPT:
-				args->print_backrefs = true;
-				break;
-			case EXTENT_OPT:
-				args->print_extents = true;
-				break;
-			default:
-				argp_usage(state);
-				parse_err = true;
-				break;
-			}
-		}
-		break;
-
-	case 'r':
-		/* Specific roots being requested- clear them all to start */
-		if (!args->roots_requested) {
-			clear_roots();
-			if (!args->allocs_requested)
-				args->walk_allocs = false;
-			args->roots_requested = true;
-		}
-
-		subopts = arg;
-		while (*subopts != '\0' && !parse_err) {
-			switch (getsubopt(&subopts, root_tokens, &value)) {
-			case LOGS_OPT:
-				args->walk_logs_root = true;
-				break;
-			case FS_OPT:
-				args->walk_fs_root = true;
-				break;
-			case SRCH_OPT:
-				args->walk_srch_root = true;
-				break;
-			default:
-				argp_usage(state);
-				parse_err = true;
-				break;
-			}
-		}
-		break;
-
 	case ARGP_KEY_ARG:
 		if (!args->meta_device)
 			args->meta_device = strdup_or_error(state, arg);
 		else
 			argp_error(state, "more than one argument given");
 		break;
-
 	case ARGP_KEY_FINI:
 		if (!args->meta_device)
 			argp_error(state, "no metadata device argument given");
-
-		/*
-		 * For backwards compatibility, translate -S. Should we warn if
-		 * this conflicts with other explicit options?
-		 */
-		if (args->skip_likely_huge) {
-			if (!args->allocs_requested)
-				args->walk_allocs = false;
-			args->walk_fs_root = false;
-			args->walk_srch_root = false;
-		}
-
 		break;
-
 	default:
 		break;
 	}
@@ -1390,10 +1192,7 @@ static int parse_opt(int key, char *arg, struct argp_state *state)
 }
 
 static struct argp_option options[] = {
-	{ "allocs", 'a', NULL, 0, "Print metadata and data alloc lists" },
-	{ "items", 'i', "ITEMS", 0, "Item(s) to print (inode, xattr, dirent, symlink, backref, extent)" },
-	{ "roots", 'r', "ROOTS", 0, "Tree root(s) to walk (logs, srch, fs)" },
-	{ "skip-likely-huge", 'S', NULL, 0, "Skip allocs, srch root and fs root to minimize output size" },
+	{ "skip-likely-huge", 'S', NULL, 0, "Skip large structures to minimize output size"},
 	{ NULL }
 };
 
@@ -1406,15 +1205,17 @@ static struct argp argp = {
 
 static int print_cmd(int argc, char **argv)
 {
+	struct print_args print_args = {NULL};
 	int ret;
 
 	ret = argp_parse(&argp, argc, argv, 0, NULL, &print_args);
 	if (ret)
 		return ret;
 
-	return do_print();
+	return do_print(&print_args);
 }
 
+
 static void __attribute__((constructor)) print_ctor(void)
 {
 	cmd_register_argp("print", &argp, GROUP_DEBUG, print_cmd);

utils/tuned/40-scoutfs.conf

@@ -0,0 +1,9 @@
+#
+# scoutfs tuned recommendation
+#
+
+# If the system has support for mounting scoutfs filesystems, which is
+# valid for client mounts and quorum mounts. We then always recommend
+# the scoutfs profile.
+[scoutfs]
+/proc/filesystems=scoutfs

utils/tuned/tuned.conf

@@ -0,0 +1,40 @@
+#
+# ScoutFS specific tuned profile
+#
+
+# The parameters below are a mix of settings present in the throughput-performance
+# profile as well as the latency-performance profile. Generally speaking, we
+# want to encourage the system to avoid swap and accumulating large amounts of
+# dirty data, as this can cause reclaim to lead to congestion.
+
+# Enable this profile with `$ sudo tuned-adm profile scoutfs`
+
+# linux default values are marked with [<value>] for reference.
+
+[main]
+summary=Optimize for production scoutfs deployment
+description=Configures the system for production scoutfs filesystem server deployment.
+
+# network-throughput sets some larger buffers useful for 40gbe deployments
+# network-throughput also inherits throughput-performance
+include=network-throughput
+
+[vm]
+# throughput-performance sets dirty_bytes to 40% (much larger than linux default), but
+# this allows the accumulation of large backlogs of writeback. We prefer to writeback
+# often and early to avoid congestion [20%]
+dirty_bytes = 10%
+# start writing back at this amount [10%]
+dirty_background_bytes = 5%
+
+[sysctl]
+# the kernel default is 60. Lower it to instruct the kernel that swapping is
+# expensive and we want to avoid it. We assume scoutfs deployments have ample
+# available RAM. [60]
+vm.swappiness = 10
+
+# increase pdflush runs so it can more aggressively write out dirty data [500]
+vm.dirty_writeback_centisecs = 300
+
+# decrease time dirty data will linger before being written back [3000]
+vm.dirty_expire_centisecs = 2000