tools/xenstore: make the internal memory data base the default

Having a file backed data base has the only advantage of being capable
to dump the contents of it while Xenstore is running, and potentially
using less swap space in case the data base can't be kept in memory.

It has the major disadvantage of a huge performance overhead: switching
to keep the data base in memory only speeds up live update of xenstored
with 120000 nodes from 20 minutes to 11 seconds. A complete tree walk
of this configuration will be reduced from 7 seconds to 280 msecs
(measured by "xenstore-control check").

So make the internal memory data base the default and enhance the
"--internal-db" command line parameter to take an optional parameter
allowing to switch the internal data base back to the file based one.

This is part of XSA-419.

Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Julien Grall <jgrall@amazon.com>
(cherry picked from commit d174fefa90487ddd25ebc618028f67b2e8a1f795)

diff --git a/tools/helpers/init-xenstore-domain.c b/tools/helpers/init-xenstore-domain.c
index 4ce8299c3c3613bb97f4133bdd699a7f700648a1..2dfc78d4de56755c197e8d963be5fe4e8b2add9c 100644 (file)
--- a/tools/helpers/init-xenstore-domain.c
+++ b/tools/helpers/init-xenstore-domain.c
@@ -137,9 +137,9 @@ static int build(xc_interface *xch)
     }
 
     if ( param )
-        snprintf(cmdline, 512, "--event %d --internal-db %s", rv, param);
+        snprintf(cmdline, 512, "--event %d %s", rv, param);
     else
-        snprintf(cmdline, 512, "--event %d --internal-db", rv);
+        snprintf(cmdline, 512, "--event %d", rv);
 
     dom = xc_dom_allocate(xch, cmdline, NULL);
     if ( !dom )

diff --git a/tools/xenstore/xenstored_core.c b/tools/xenstore/xenstored_core.c
index eb1d7c2b3b9bc016f274c90f2d186e98a599c4a9..7ed852fd79586cbc894ff4a2a5a9fe3e4ab8ca30 100644 (file)
--- a/tools/xenstore/xenstored_core.c
+++ b/tools/xenstore/xenstored_core.c
@@ -2134,7 +2134,7 @@ static void accept_connection(int sock, bool canwrite)
 }
 #endif
 
-static int tdb_flags;
+static int tdb_flags = TDB_INTERNAL | TDB_NOLOCK;
 
 /* We create initial nodes manually. */
 static void manual_node(const char *name, const char *child)
@@ -2446,7 +2446,8 @@ static void usage(void)
 "                          watch-event: time a watch-event is kept pending\n"
 "  -R, --no-recovery       to request that no recovery should be attempted when\n"
 "                          the store is corrupted (debug only),\n"
-"  -I, --internal-db       store database in memory, not on disk\n"
+"  -I, --internal-db [on|off] store database in memory, not on disk, default is\n"
+"                          memory, with \"--internal-db off\" it is on disk\n"
 "  -K, --keep-orphans      don't delete nodes owned by a domain when the\n"
 "                          domain is deleted (this is a security risk!)\n"
 "  -V, --verbose           to request verbose execution.\n");
@@ -2471,7 +2472,7 @@ static struct option options[] = {
        { "quota-soft", 1, NULL, 'q' },
        { "timeout", 1, NULL, 'w' },
        { "no-recovery", 0, NULL, 'R' },
-       { "internal-db", 0, NULL, 'I' },
+       { "internal-db", 2, NULL, 'I' },
        { "keep-orphans", 0, NULL, 'K' },
        { "verbose", 0, NULL, 'V' },
        { "watch-nb", 1, NULL, 'W' },
@@ -2547,7 +2548,8 @@ int main(int argc, char *argv[])
        int timeout;
 
 
-       while ((opt = getopt_long(argc, argv, "DE:F:HKNPS:t:A:Q:q:T:RVW:w:", options,
+       while ((opt = getopt_long(argc, argv,
+                                 "DE:F:HI::KNPS:t:A:Q:q:T:RVW:w:", options,
                                  NULL)) != -1) {
                switch (opt) {
                case 'D':
@@ -2581,7 +2583,8 @@ int main(int argc, char *argv[])
                        tracefile = optarg;
                        break;
                case 'I':
-                       tdb_flags = TDB_INTERNAL|TDB_NOLOCK;
+                       if (optarg && !strcmp(optarg, "off"))
+                               tdb_flags = 0;
                        break;
                case 'K':
                        keep_orphans = true;