Merge pull request ceph#3979 from ceph/wip-rbd-map-options

rbd: default map options, new options, misc

Reviewed-by: Josh Durgin <jdurgin@redhat.com>

doc/man/8/rbd.rst

@@ -366,6 +366,17 @@ the running kernel.
 
 * nocrc - Disable CRC32C checksumming for data writes.
 
+* cephx_require_signatures - Require cephx message signing, i.e. MSG_AUTH
+  feature bit (since 3.19, default).
+
+* nocephx_require_signatures - Don't require cephx message signing (since
+  3.19).
+
+* tcp_nodelay - Disable Nagle's algorithm on client sockets (since 4.0,
+  default).
+
+* notcp_nodelay - Enable Nagle's algorithm on client sockets (since 4.0).
+
 * osdkeepalive=x - OSD keepalive timeout (default is 5 seconds).
 
 * osd_idle_ttl=x - OSD idle TTL (default is 60 seconds).

man/rbd.8

@@ -1,6 +1,6 @@
 .\" Man page generated from reStructuredText.
 .
-.TH "RBD" "8" "January 12, 2014" "dev" "Ceph"
+.TH "RBD" "8" "March 12, 2015" "dev" "Ceph"
 .SH NAME
 rbd \- manage rados block device (RBD) images
 .
@@ -187,6 +187,25 @@ See map options section below for more details.
 .B \-\-read\-only
 Map the image read\-only.  Equivalent to \-o ro.
 .UNINDENT
+.INDENT 0.0
+.TP
+.B \-\-image\-features features
+Specifies which RBD format 2 features are to be enabled when creating
+an image. The numbers from the desired features below should be added
+to compute the parameter value:
+.sp
++1: layering support
++2: striping v2 support
++4: exclusive locking support
++8: object map support
+.UNINDENT
+.INDENT 0.0
+.TP
+.B \-\-image\-shared
+Specifies that the image will be used concurrently by multiple clients.
+This will disable features that are dependent upon exclusive ownership
+of the image.
+.UNINDENT
 .SH COMMANDS
 .INDENT 0.0
 .TP
@@ -207,7 +226,8 @@ Will create a new rbd image. You must also specify the size via \-\-size.  The
 .B \fBclone\fP [\fIparent\-snapname\fP] [\fIimage\-name\fP]
 Will create a clone (copy\-on\-write child) of the parent snapshot.
 Object order will be identical to that of the parent image unless
-specified. Size will be the same as the parent snapshot.
+specified. Size will be the same as the parent snapshot. The \-\-stripe\-unit
+and \-\-stripe\-count arguments are optional, but must be used together.
 .sp
 The parent snapshot must be protected (see \fIrbd snap protect\fP).
 This requires image format 2.
@@ -242,6 +262,9 @@ Creates a new image and imports its data from path (use \- for
 stdin).  The import operation will try to create sparse rbd images
 if possible.  For import from stdin, the sparsification unit is
 the data block size of the destination image (1 << order).
+.sp
+The \-\-stripe\-unit and \-\-stripe\-count arguments are optional, but must be
+used together.
 .TP
 .B \fBexport\-diff\fP [\fIimage\-name\fP] [\fIdest\-path\fP] [\-\-from\-snap \fIsnapname\fP]
 Exports an incremental diff for an image to dest path (use \- for stdout).  If
@@ -251,6 +274,14 @@ using the standard \-\-snap option or @snap syntax (see below).  The image diff
 metadata about image size changes, and the start and end snapshots.  It efficiently represents
 discarded or \(aqzero\(aq regions of the image.
 .TP
+.B \fBmerge\-diff\fP [\fIfirst\-diff\-path\fP] [\fIsecond\-diff\-path\fP] [\fImerged\-diff\-path\fP]
+Merge two continuous incremental diffs of an image into one single diff. The
+first diff\(aqs end snapshot must be equal with the second diff\(aqs start snapshot.
+The first diff could be \- for stdin, and merged diff could be \- for stdout, which
+enables multiple diff files to be merged using something like
+\(aqrbd merge\-diff first second \- | rbd merge\-diff \- third result\(aq. Note this command
+currently only support the source incremental diff with stripe_count == 1
+.TP
 .B \fBimport\-diff\fP [\fIsrc\-path\fP] [\fIimage\-name\fP]
 Imports an incremental diff of an image and applies it to the current image.  If the diff
 was generated relative to a start snapshot, we verify that snapshot already exists before
@@ -311,6 +342,9 @@ Unmaps the block device that was mapped via the rbd kernel module.
 .B \fBshowmapped\fP
 Show the rbd images that are mapped via the rbd kernel module.
 .TP
+.B \fBstatus\fP [\fIimage\-name\fP]
+Show the status of the image, including which clients have it open.
+.TP
 .B \fBlock\fP list [\fIimage\-name\fP]
 Show locks held on the image. The first column is the locker
 to use with the \fIlock remove\fP command.
@@ -361,28 +395,21 @@ The striping is controlled by three parameters:
 .INDENT 0.0
 .TP
 .B order
-.TP
-.B The size of objects we stripe over is a power of two, specifially 2^[*order*] bytes.  The default
-.TP
-.B is 22, or 4 MB.
+The size of objects we stripe over is a power of two, specifically 2^[\fIorder\fP] bytes.  The default
+is 22, or 4 MB.
 .UNINDENT
 .INDENT 0.0
 .TP
 .B stripe_unit
-.TP
-.B Each [*stripe_unit*] contiguous bytes are stored adjacently in the same object, before we move on
-.TP
-.B to the next object.
+Each [\fIstripe_unit\fP] contiguous bytes are stored adjacently in the same object, before we move on
+to the next object.
 .UNINDENT
 .INDENT 0.0
 .TP
 .B stripe_count
-.TP
-.B After we write [*stripe_unit*] bytes to [*stripe_count*] objects, we loop back to the initial object
-.TP
-.B and write another stripe, until the object reaches its maximum size (as specified by [*order*].  At that
-.TP
-.B point, we move on to the next [*stripe_count*] objects.
+After we write [\fIstripe_unit\fP] bytes to [\fIstripe_count\fP] objects, we loop back to the initial object
+and write another stripe, until the object reaches its maximum size (as specified by [\fIorder\fP].  At that
+point, we move on to the next [\fIstripe_count\fP] objects.
 .UNINDENT
 .sp
 By default, [\fIstripe_unit\fP] is the same as the object size and [\fIstripe_count\fP] is 1.  Specifying a different
@@ -408,6 +435,17 @@ crc \- Enable CRC32C checksumming for data writes (default).
 .IP \(bu 2
 nocrc \- Disable CRC32C checksumming for data writes.
 .IP \(bu 2
+cephx_require_signatures \- Require cephx message signing, i.e. MSG_AUTH
+feature bit (since 3.19, default).
+.IP \(bu 2
+nocephx_require_signatures \- Don\(aqt require cephx message signing (since
+3.19).
+.IP \(bu 2
+tcp_nodelay \- Disable Nagle\(aqs algorithm on client sockets (since 4.0,
+default).
+.IP \(bu 2
+notcp_nodelay \- Enable Nagle\(aqs algorithm on client sockets (since 4.0).
+.IP \(bu 2
 osdkeepalive=x \- OSD keepalive timeout (default is 5 seconds).
 .IP \(bu 2
 osd_idle_ttl=x \- OSD idle TTL (default is 60 seconds).
@@ -604,7 +642,7 @@ rbd lock remove mypool/myimage mylockid client.2485
 .UNINDENT
 .SH AVAILABILITY
 .sp
-\fBrbd\fP is part of Ceph, a massively scalable, open-source, distributed storage system. Please refer to
+\fBrbd\fP is part of Ceph, a massively scalable, open\-source, distributed storage system. Please refer to
 the Ceph documentation at \fI\%http://ceph.com/docs\fP for more information.
 .SH SEE ALSO
 .sp

src/common/config_opts.h

@@ -900,6 +900,8 @@ OPTION(rbd_default_features, OPT_INT, 3) // only applies to format 2 images
 					 // +1 for layering, +2 for stripingv2,
 					 // +4 for exclusive lock, +8 for object map
 
+OPTION(rbd_default_map_options, OPT_STR, "") // default rbd map -o / --options
+
 OPTION(nss_db_path, OPT_STR, "") // path to nss db
 
 

src/rbd.cc

@@ -2267,18 +2267,19 @@ static int do_kernel_map(const char *poolname, const char *imgname,
   if (r < 0)
     return r;
 
-  for (map<string, string>::const_iterator it = map_options.begin();
-       it != map_options.end();
-       ++it) {
+  for (map<string, string>::iterator it = map_options.begin();
+       it != map_options.end(); ) {
     // for compatibility with < 3.7 kernels, assume that rw is on by
     // default and omit it even if it was specified by the user
     // (see ceph.git commit fb0f1986449b)
-    if (it->first == "rw" && it->second == "rw")
-      continue;
-
-    if (it != map_options.begin())
-      oss << ",";
-    oss << it->second;
+    if (it->first == "rw" && it->second == "rw") {
+      map_options.erase(it++);
+    } else {
+      if (it != map_options.begin())
+        oss << ",";
+      oss << it->second;
+      ++it;
+    }
   }
 
   r = krbd_map(krbd, poolname, imgname, snapname, oss.str().c_str(), &devnode);
@@ -2356,11 +2357,6 @@ static string map_option_int_cb(const char *value_char)
 
 static void put_map_option(const string key, string val)
 {
-  map<string, string>::const_iterator it = map_options.find(key);
-  if (it != map_options.end()) {
-    cerr << "rbd: warning: redefining map option " << key << ": '"
-         << it->second << "' -> '" << val << "'" << std::endl;
-  }
   map_options[key] = val;
 }
 
@@ -2403,6 +2399,12 @@ static int parse_map_options(char *options)
       put_map_option("share", this_char);
     } else if (!strcmp(this_char, "crc") || !strcmp(this_char, "nocrc")) {
       put_map_option("crc", this_char);
+    } else if (!strcmp(this_char, "cephx_require_signatures") ||
+               !strcmp(this_char, "nocephx_require_signatures")) {
+      put_map_option("cephx_require_signatures", this_char);
+    } else if (!strcmp(this_char, "tcp_nodelay") ||
+               !strcmp(this_char, "notcp_nodelay")) {
+      put_map_option("tcp_nodelay", this_char);
     } else if (!strcmp(this_char, "mount_timeout")) {
       if (put_map_option_value("mount_timeout", value_char, map_option_int_cb))
         return 1;
@@ -2595,6 +2597,7 @@ int main(int argc, const char **argv)
     *lock_tag = NULL, *output_format = "plain",
     *fromsnapname = NULL,
     *first_diff = NULL, *second_diff = NULL;
+  char *cli_map_options = NULL;
   bool lflag = false;
   int pretty_format = 0;
   long long stripe_unit = 0, stripe_count = 0;
@@ -2679,11 +2682,7 @@ int main(int argc, const char **argv)
     } else if (ceph_argparse_flag(args, i, "--no-settle", (char *)NULL)) {
       cerr << "rbd: --no-settle is deprecated" << std::endl;
     } else if (ceph_argparse_witharg(args, i, &val, "-o", "--options", (char*)NULL)) {
-      char *map_options = strdup(val.c_str());
-      if (parse_map_options(map_options)) {
-        cerr << "rbd: couldn't parse map options" << std::endl;
-        return EXIT_FAILURE;
-      }
+      cli_map_options = strdup(val.c_str());
     } else if (ceph_argparse_flag(args, i, "--read-only", (char *)NULL)) {
       // --read-only is equivalent to -o ro
       put_map_option("rw", "ro");
@@ -2915,6 +2914,20 @@ if (!set_conf_param(v, p1, p2, p3)) { \
     return EXIT_FAILURE;
   }
 
+  if (opt_cmd == OPT_MAP) {
+    char *default_map_options = strdup(g_conf->rbd_default_map_options.c_str());
+
+    // parse default options first so they can be overwritten by cli options
+    if (parse_map_options(default_map_options)) {
+      cerr << "rbd: couldn't parse default map options" << std::endl;
+      return EXIT_FAILURE;
+    }
+    if (cli_map_options && parse_map_options(cli_map_options)) {
+      cerr << "rbd: couldn't parse map options" << std::endl;
+      return EXIT_FAILURE;
+    }
+  }
+
   if (opt_cmd == OPT_UNMAP && !devpath) {
     cerr << "rbd: device path was not specified" << std::endl;
     return EXIT_FAILURE;