All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH 2/2] rbd: add an option for md5 checksumming (v4)
@ 2011-08-26 20:22 Christian Brunner
  2011-08-27 19:41 ` H. Andres Lagar-Cavilla
  0 siblings, 1 reply; 2+ messages in thread
From: Christian Brunner @ 2011-08-26 20:22 UTC (permalink / raw)
  To: ceph-devel

We needed to get an md5 checksum of an rbd image. Since librbd is using a
lot of sparse operations, this was not possible without writing an image
to a local disk.

With this patch exporting the image is no longer needed. You can do
"rbd md5 image" and you will get the same output as you would call "md5sum".

V1 -> V2:
- use ceph::crypto classes
- support for sha1
- skip holes

V2 -> V3:
- better handling for holes

V3 -> V4:
- hopefully the hole handling is done right now...

---
 src/rbd.cc |  129 +++++++++++++++++++++++++++++++++++++++++++++++++----------
 1 files changed, 107 insertions(+), 22 deletions(-)

diff --git a/src/rbd.cc b/src/rbd.cc
index a18c029..8dfbb37 100644
--- a/src/rbd.cc
+++ b/src/rbd.cc
@@ -38,6 +38,8 @@
 #include <time.h>
 #include <sys/ioctl.h>
 
+#include "common/ceph_crypto.h"
+
 #include "include/rbd_types.h"
 
 #include <linux/fs.h>
@@ -50,6 +52,29 @@
 static string dir_oid = RBD_DIRECTORY;
 static string dir_info_oid = RBD_INFO;
 
+enum {
+  OPT_NO_CMD = 0,
+  OPT_LIST,
+  OPT_INFO,
+  OPT_CREATE,
+  OPT_RESIZE,
+  OPT_RM,
+  OPT_EXPORT,
+  OPT_IMPORT,
+  OPT_COPY,
+  OPT_RENAME,
+  OPT_SNAP_CREATE,
+  OPT_SNAP_ROLLBACK,
+  OPT_SNAP_REMOVE,
+  OPT_SNAP_LIST,
+  OPT_WATCH,
+  OPT_MAP,
+  OPT_UNMAP,
+  OPT_SHOWMAPPED,
+  OPT_MD5,
+  OPT_SHA1,
+};
+
 void usage()
 {
   cout << "usage: rbd [-n <auth user>] [OPTIONS] <cmd> ...\n"
@@ -65,6 +90,8 @@ void usage()
        << "  export [image-name] [dest-path]           export image to file\n"
        << "  import [path] [dst-image]                 import image from file (dest defaults\n"
        << "                                            as the filename part of file)\n"
+       << "  md5 [image-name]                          print md5 checksum for image\n"
+       << "  sha1 [image-name]                         print sha1 checksum for image\n"
        << "  <cp | copy> [src-image] [dest-image]      copy image to dest\n"
        << "  <mv | rename> [src-image] [dest-image]    copy image to dest\n"
        << "  snap ls [image-name]                      dump list of image snapshots\n"
@@ -262,6 +289,68 @@ static int do_export(librbd::Image& image, const char *path)
   return 0;
 }
 
+static int hash_read_cb(uint64_t ofs, size_t len, const char *buf, void *arg)
+{
+  ceph::crypto::Digest *Hash = (ceph::crypto::Digest *)arg;
+  char *tempbuf = NULL;
+
+  if (!buf) {
+    tempbuf = (char *) calloc(len, sizeof(char));
+    if (!tempbuf)
+      return -ENOMEM;
+    buf = tempbuf;
+  }
+  Hash->Update((const byte *) buf, len);
+
+  free(tempbuf);
+
+  return 0;
+}
+
+static int do_hash(librbd::Image& image, const char *imgname, int opt_cmd)
+{
+  int64_t r, i, digest_size;
+  byte *digest;
+  char hexval[] = "0123456789abcdef";
+  char *hexdigest;
+  librbd::image_info_t info;
+  ceph::crypto::Digest *Hash;
+
+  if (opt_cmd == OPT_MD5) {
+    Hash = (ceph::crypto::Digest *) new ceph::crypto::MD5;
+  } else if (opt_cmd == OPT_SHA1) {
+    Hash = (ceph::crypto::Digest *) new ceph::crypto::SHA1;
+  } else {
+    return -1;
+  }
+
+  r = image.stat(info, sizeof(info));
+  if (r < 0)
+    return r;
+
+  r = image.read_iterate(0, info.size, hash_read_cb, (void *)Hash);
+  if (r < 0)
+    return r;
+
+  digest_size = Hash->DigestSize();
+  digest = (byte *) malloc(digest_size);
+  hexdigest = (char *) malloc((digest_size * 2 + 1) * sizeof(char));
+  Hash->Final(digest);
+
+  for(i = 0; i < digest_size; i++){
+    hexdigest[i*2] = hexval[((digest[i] >> 4) & 0xF)];
+    hexdigest[(i*2) + 1] = hexval[(digest[i]) & 0x0F];
+  }
+  hexdigest[(digest_size*2)] = '\0';
+
+  cout << hexdigest << "  " << imgname << std::endl;
+
+  free(hexdigest);
+  free(digest);
+
+  return 0;
+}
+
 static const char *imgname_from_path(const char *path)
 {
   const char *imgname;
@@ -720,27 +809,6 @@ static int do_kernel_rm(const char *dev)
   return r;
 }
 
-enum {
-  OPT_NO_CMD = 0,
-  OPT_LIST,
-  OPT_INFO,
-  OPT_CREATE,
-  OPT_RESIZE,
-  OPT_RM,
-  OPT_EXPORT,
-  OPT_IMPORT,
-  OPT_COPY,
-  OPT_RENAME,
-  OPT_SNAP_CREATE,
-  OPT_SNAP_ROLLBACK,
-  OPT_SNAP_REMOVE,
-  OPT_SNAP_LIST,
-  OPT_WATCH,
-  OPT_MAP,
-  OPT_UNMAP,
-  OPT_SHOWMAPPED,
-};
-
 static int get_cmd(const char *cmd, bool *snapcmd)
 {
   if (strcmp(cmd, "snap") == 0) {
@@ -766,6 +834,10 @@ static int get_cmd(const char *cmd, bool *snapcmd)
       return OPT_EXPORT;
     if (strcmp(cmd, "import") == 0)
       return OPT_IMPORT;
+    if (strcmp(cmd, "md5") == 0)
+      return OPT_MD5;
+    if (strcmp(cmd, "sha1") == 0)
+      return OPT_SHA1;
     if (strcmp(cmd, "copy") == 0 ||
         strcmp(cmd, "cp") == 0)
       return OPT_COPY;
@@ -888,6 +960,10 @@ int main(int argc, const char **argv)
           case OPT_IMPORT:
             set_conf_param(CEPH_ARGPARSE_VAL, &path, &destname);
             break;
+          case OPT_MD5:
+          case OPT_SHA1:
+            set_conf_param(CEPH_ARGPARSE_VAL, &imgname, NULL);
+            break;
           case OPT_COPY:
           case OPT_RENAME:
             set_conf_param(CEPH_ARGPARSE_VAL, &imgname, &destname);
@@ -966,7 +1042,7 @@ int main(int argc, const char **argv)
       (opt_cmd == OPT_RESIZE || opt_cmd == OPT_INFO || opt_cmd == OPT_SNAP_LIST ||
        opt_cmd == OPT_SNAP_CREATE || opt_cmd == OPT_SNAP_ROLLBACK ||
        opt_cmd == OPT_SNAP_REMOVE || opt_cmd == OPT_EXPORT || opt_cmd == OPT_WATCH ||
-       opt_cmd == OPT_COPY)) {
+       opt_cmd == OPT_COPY || opt_cmd == OPT_MD5 || opt_cmd == OPT_SHA1 )) {
     r = rbd.open(io_ctx, image, imgname);
     if (r < 0) {
       cerr << "error opening image " << imgname << ": " << strerror(r) << std::endl;
@@ -1127,6 +1203,15 @@ int main(int argc, const char **argv)
     }
     break;
 
+  case OPT_MD5:
+  case OPT_SHA1:
+    r = do_hash(image, imgname, opt_cmd);
+    if (r < 0) {
+      cerr << "hashing failed: " << strerror(-r) << std::endl;
+      exit(1);
+    }
+    break;
+
   case OPT_COPY:
     r = do_copy(image, dest_io_ctx, destname);
     if (r < 0) {
-- 
1.7.1


^ permalink raw reply related	[flat|nested] 2+ messages in thread

* Re: [PATCH 2/2] rbd: add an option for md5 checksumming (v4)
  2011-08-26 20:22 [PATCH 2/2] rbd: add an option for md5 checksumming (v4) Christian Brunner
@ 2011-08-27 19:41 ` H. Andres Lagar-Cavilla
  0 siblings, 0 replies; 2+ messages in thread
From: H. Andres Lagar-Cavilla @ 2011-08-27 19:41 UTC (permalink / raw)
  To: Christian Brunner; +Cc: ceph-devel

I know this has been an ongoing and very well informed discussion. Here's my two cents

#define SIZE some_multiple_of_4kb
fd = open("/dev/zero", O_RDONLY)
zerobuf = mmap(0, SIZE, PROT_READ, MAP_SHARED, fd, 0);

Two syscalls at the beginning of the routine and you have a zero buffer of size SIZE. It continually maps the same single page of all zeros (kernel zero_page). The only space overhead is page table entries, negligible. Reuse it at will every time you need to feed zeros to md5. No runtime overhead of any kind, O(1).

Andres

On Aug 26, 2011, at 4:22 PM, Christian Brunner wrote:

> We needed to get an md5 checksum of an rbd image. Since librbd is using a
> lot of sparse operations, this was not possible without writing an image
> to a local disk.
> 
> With this patch exporting the image is no longer needed. You can do
> "rbd md5 image" and you will get the same output as you would call "md5sum".
> 
> V1 -> V2:
> - use ceph::crypto classes
> - support for sha1
> - skip holes
> 
> V2 -> V3:
> - better handling for holes
> 
> V3 -> V4:
> - hopefully the hole handling is done right now...
> 
> ---
> src/rbd.cc |  129 +++++++++++++++++++++++++++++++++++++++++++++++++----------
> 1 files changed, 107 insertions(+), 22 deletions(-)
> 
> diff --git a/src/rbd.cc b/src/rbd.cc
> index a18c029..8dfbb37 100644
> --- a/src/rbd.cc
> +++ b/src/rbd.cc
> @@ -38,6 +38,8 @@
> #include <time.h>
> #include <sys/ioctl.h>
> 
> +#include "common/ceph_crypto.h"
> +
> #include "include/rbd_types.h"
> 
> #include <linux/fs.h>
> @@ -50,6 +52,29 @@
> static string dir_oid = RBD_DIRECTORY;
> static string dir_info_oid = RBD_INFO;
> 
> +enum {
> +  OPT_NO_CMD = 0,
> +  OPT_LIST,
> +  OPT_INFO,
> +  OPT_CREATE,
> +  OPT_RESIZE,
> +  OPT_RM,
> +  OPT_EXPORT,
> +  OPT_IMPORT,
> +  OPT_COPY,
> +  OPT_RENAME,
> +  OPT_SNAP_CREATE,
> +  OPT_SNAP_ROLLBACK,
> +  OPT_SNAP_REMOVE,
> +  OPT_SNAP_LIST,
> +  OPT_WATCH,
> +  OPT_MAP,
> +  OPT_UNMAP,
> +  OPT_SHOWMAPPED,
> +  OPT_MD5,
> +  OPT_SHA1,
> +};
> +
> void usage()
> {
>   cout << "usage: rbd [-n <auth user>] [OPTIONS] <cmd> ...\n"
> @@ -65,6 +90,8 @@ void usage()
>        << "  export [image-name] [dest-path]           export image to file\n"
>        << "  import [path] [dst-image]                 import image from file (dest defaults\n"
>        << "                                            as the filename part of file)\n"
> +       << "  md5 [image-name]                          print md5 checksum for image\n"
> +       << "  sha1 [image-name]                         print sha1 checksum for image\n"
>        << "  <cp | copy> [src-image] [dest-image]      copy image to dest\n"
>        << "  <mv | rename> [src-image] [dest-image]    copy image to dest\n"
>        << "  snap ls [image-name]                      dump list of image snapshots\n"
> @@ -262,6 +289,68 @@ static int do_export(librbd::Image& image, const char *path)
>   return 0;
> }
> 
> +static int hash_read_cb(uint64_t ofs, size_t len, const char *buf, void *arg)
> +{
> +  ceph::crypto::Digest *Hash = (ceph::crypto::Digest *)arg;
> +  char *tempbuf = NULL;
> +
> +  if (!buf) {
> +    tempbuf = (char *) calloc(len, sizeof(char));
> +    if (!tempbuf)
> +      return -ENOMEM;
> +    buf = tempbuf;
> +  }
> +  Hash->Update((const byte *) buf, len);
> +
> +  free(tempbuf);
> +
> +  return 0;
> +}
> +
> +static int do_hash(librbd::Image& image, const char *imgname, int opt_cmd)
> +{
> +  int64_t r, i, digest_size;
> +  byte *digest;
> +  char hexval[] = "0123456789abcdef";
> +  char *hexdigest;
> +  librbd::image_info_t info;
> +  ceph::crypto::Digest *Hash;
> +
> +  if (opt_cmd == OPT_MD5) {
> +    Hash = (ceph::crypto::Digest *) new ceph::crypto::MD5;
> +  } else if (opt_cmd == OPT_SHA1) {
> +    Hash = (ceph::crypto::Digest *) new ceph::crypto::SHA1;
> +  } else {
> +    return -1;
> +  }
> +
> +  r = image.stat(info, sizeof(info));
> +  if (r < 0)
> +    return r;
> +
> +  r = image.read_iterate(0, info.size, hash_read_cb, (void *)Hash);
> +  if (r < 0)
> +    return r;
> +
> +  digest_size = Hash->DigestSize();
> +  digest = (byte *) malloc(digest_size);
> +  hexdigest = (char *) malloc((digest_size * 2 + 1) * sizeof(char));
> +  Hash->Final(digest);
> +
> +  for(i = 0; i < digest_size; i++){
> +    hexdigest[i*2] = hexval[((digest[i] >> 4) & 0xF)];
> +    hexdigest[(i*2) + 1] = hexval[(digest[i]) & 0x0F];
> +  }
> +  hexdigest[(digest_size*2)] = '\0';
> +
> +  cout << hexdigest << "  " << imgname << std::endl;
> +
> +  free(hexdigest);
> +  free(digest);
> +
> +  return 0;
> +}
> +
> static const char *imgname_from_path(const char *path)
> {
>   const char *imgname;
> @@ -720,27 +809,6 @@ static int do_kernel_rm(const char *dev)
>   return r;
> }
> 
> -enum {
> -  OPT_NO_CMD = 0,
> -  OPT_LIST,
> -  OPT_INFO,
> -  OPT_CREATE,
> -  OPT_RESIZE,
> -  OPT_RM,
> -  OPT_EXPORT,
> -  OPT_IMPORT,
> -  OPT_COPY,
> -  OPT_RENAME,
> -  OPT_SNAP_CREATE,
> -  OPT_SNAP_ROLLBACK,
> -  OPT_SNAP_REMOVE,
> -  OPT_SNAP_LIST,
> -  OPT_WATCH,
> -  OPT_MAP,
> -  OPT_UNMAP,
> -  OPT_SHOWMAPPED,
> -};
> -
> static int get_cmd(const char *cmd, bool *snapcmd)
> {
>   if (strcmp(cmd, "snap") == 0) {
> @@ -766,6 +834,10 @@ static int get_cmd(const char *cmd, bool *snapcmd)
>       return OPT_EXPORT;
>     if (strcmp(cmd, "import") == 0)
>       return OPT_IMPORT;
> +    if (strcmp(cmd, "md5") == 0)
> +      return OPT_MD5;
> +    if (strcmp(cmd, "sha1") == 0)
> +      return OPT_SHA1;
>     if (strcmp(cmd, "copy") == 0 ||
>         strcmp(cmd, "cp") == 0)
>       return OPT_COPY;
> @@ -888,6 +960,10 @@ int main(int argc, const char **argv)
>           case OPT_IMPORT:
>             set_conf_param(CEPH_ARGPARSE_VAL, &path, &destname);
>             break;
> +          case OPT_MD5:
> +          case OPT_SHA1:
> +            set_conf_param(CEPH_ARGPARSE_VAL, &imgname, NULL);
> +            break;
>           case OPT_COPY:
>           case OPT_RENAME:
>             set_conf_param(CEPH_ARGPARSE_VAL, &imgname, &destname);
> @@ -966,7 +1042,7 @@ int main(int argc, const char **argv)
>       (opt_cmd == OPT_RESIZE || opt_cmd == OPT_INFO || opt_cmd == OPT_SNAP_LIST ||
>        opt_cmd == OPT_SNAP_CREATE || opt_cmd == OPT_SNAP_ROLLBACK ||
>        opt_cmd == OPT_SNAP_REMOVE || opt_cmd == OPT_EXPORT || opt_cmd == OPT_WATCH ||
> -       opt_cmd == OPT_COPY)) {
> +       opt_cmd == OPT_COPY || opt_cmd == OPT_MD5 || opt_cmd == OPT_SHA1 )) {
>     r = rbd.open(io_ctx, image, imgname);
>     if (r < 0) {
>       cerr << "error opening image " << imgname << ": " << strerror(r) << std::endl;
> @@ -1127,6 +1203,15 @@ int main(int argc, const char **argv)
>     }
>     break;
> 
> +  case OPT_MD5:
> +  case OPT_SHA1:
> +    r = do_hash(image, imgname, opt_cmd);
> +    if (r < 0) {
> +      cerr << "hashing failed: " << strerror(-r) << std::endl;
> +      exit(1);
> +    }
> +    break;
> +
>   case OPT_COPY:
>     r = do_copy(image, dest_io_ctx, destname);
>     if (r < 0) {
> -- 
> 1.7.1
> 
> --
> To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html


^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2011-08-27 19:32 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2011-08-26 20:22 [PATCH 2/2] rbd: add an option for md5 checksumming (v4) Christian Brunner
2011-08-27 19:41 ` H. Andres Lagar-Cavilla

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.