diff mbox

[v20,3/5] block/gluster: deprecate rdma support

Message ID 1468947453-5433-4-git-send-email-prasanna.kalever@redhat.com (mailing list archive)
State New, archived
Headers show

Commit Message

Prasanna Kumar Kalever July 19, 2016, 4:57 p.m. UTC
gluster volfile server fetch happens through unix and/or tcp, it doesn't
support volfile fetch over rdma, the rdma code may actually mislead,
to make sure things do not break, for now we fallback to tcp when requested
for rdma with a warning.

If you are wondering how this worked all these days, its the gluster libgfapi
code which handles anything other than unix transport as socket/tcp, sad but
true.

Also gluster doesn't support ipv6 addresses, removing the ipv6 related
comments/docs section

Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>
---
 block/gluster.c | 20 ++++++++++++--------
 1 file changed, 12 insertions(+), 8 deletions(-)

Comments

Markus Armbruster July 19, 2016, 5:17 p.m. UTC | #1
Prasanna Kumar Kalever <prasanna.kalever@redhat.com> writes:

> gluster volfile server fetch happens through unix and/or tcp, it doesn't
> support volfile fetch over rdma, the rdma code may actually mislead,
> to make sure things do not break, for now we fallback to tcp when requested
> for rdma with a warning.
>
> If you are wondering how this worked all these days, its the gluster libgfapi
> code which handles anything other than unix transport as socket/tcp, sad but
> true.
>
> Also gluster doesn't support ipv6 addresses, removing the ipv6 related
> comments/docs section
>
> Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>
> ---
>  block/gluster.c | 20 ++++++++++++--------
>  1 file changed, 12 insertions(+), 8 deletions(-)
>
> diff --git a/block/gluster.c b/block/gluster.c
> index 40ee852..8a54ad4 100644
> --- a/block/gluster.c
> +++ b/block/gluster.c
> @@ -12,6 +12,7 @@
>  #include "block/block_int.h"
>  #include "qapi/error.h"
>  #include "qemu/uri.h"
> +#include "qemu/error-report.h"
>  
>  #define GLUSTER_OPT_FILENAME        "filename"
>  #define GLUSTER_OPT_DEBUG           "debug"
> @@ -134,12 +135,10 @@ static int parse_volume_options(GlusterConf *gconf, char *path)
>   *
>   * 'transport' specifies the transport type used to connect to gluster
>   * management daemon (glusterd). Valid transport types are
> - * tcp, unix and rdma. If a transport type isn't specified, then tcp
> - * type is assumed.
> + * tcp, unix. If a transport type isn't specified, then tcp type is assumed.
>   *
>   * 'host' specifies the host where the volume file specification for
> - * the given volume resides. This can be either hostname, ipv4 address
> - * or ipv6 address. ipv6 address needs to be within square brackets [ ].
> + * the given volume resides. This can be either hostname, ipv4 address.

"either hostname or ipv4 address"

Could be touched up on commit, or in a cleanup patch on top.

>   * If transport type is 'unix', then 'host' field should not be specified.
>   * The 'socket' field needs to be populated with the path to unix domain
>   * socket.
> @@ -158,11 +157,8 @@ static int parse_volume_options(GlusterConf *gconf, char *path)
>   * file=gluster://1.2.3.4/testvol/a.img
>   * file=gluster+tcp://1.2.3.4/testvol/a.img
>   * file=gluster+tcp://1.2.3.4:24007/testvol/dir/a.img
> - * file=gluster+tcp://[1:2:3:4:5:6:7:8]/testvol/dir/a.img
> - * file=gluster+tcp://[1:2:3:4:5:6:7:8]:24007/testvol/dir/a.img
>   * file=gluster+tcp://host.domain.com:24007/testvol/dir/a.img
>   * file=gluster+unix:///testvol/dir/a.img?socket=/tmp/glusterd.socket
> - * file=gluster+rdma://1.2.3.4:24007/testvol/a.img
>   */
>  static int qemu_gluster_parseuri(GlusterConf *gconf, const char *filename)
>  {
> @@ -185,7 +181,9 @@ static int qemu_gluster_parseuri(GlusterConf *gconf, const char *filename)
>          gconf->transport = g_strdup("unix");
>          is_unix = true;
>      } else if (!strcmp(uri->scheme, "gluster+rdma")) {
> -        gconf->transport = g_strdup("rdma");
> +        gconf->transport = g_strdup("tcp");
> +        error_report("Warning: rdma feature is not supported falling "
> +                     "back to tcp");

"not supported, falling back"

Could be touched up on commit, or in a cleanup patch on top.

>      } else {
>          ret = -EINVAL;
>          goto out;
> @@ -1048,6 +1046,12 @@ static BlockDriver bdrv_gluster_unix = {
>      .create_opts                  = &qemu_gluster_create_opts,
>  };
>  
> +/* rdma is deprecated (actually never supported for volfile fetch)
> + * lets maintain for the protocol compatibility, to make sure things
> + * won't break immediately for now gluster+rdma will fall back to gluster+tcp
> + * protocol with Warning
> + * TODO: remove gluster+rdma interface support
> + */
>  static BlockDriver bdrv_gluster_rdma = {
>      .format_name                  = "gluster",
>      .protocol_name                = "gluster+rdma",

This comment could use some polish, but we can do that on top.

Reviewed-by: Markus Armbruster <armbru@redhat.com>
Eric Blake July 19, 2016, 6:09 p.m. UTC | #2
On 07/19/2016 10:57 AM, Prasanna Kumar Kalever wrote:

Grammar suggestions, can be squashed in by the maintainer (in addition
to what Markus pointed out):

> gluster volfile server fetch happens through unix and/or tcp, it doesn't
> support volfile fetch over rdma, the rdma code may actually mislead,

s/rdma, the/rdma. The/

> to make sure things do not break, for now we fallback to tcp when requested
s/to make/so to make/

> for rdma with a warning.

s/rdma with/rdma, with/

> 
> If you are wondering how this worked all these days, its the gluster libgfapi
> code which handles anything other than unix transport as socket/tcp, sad but
> true.
> 
> Also gluster doesn't support ipv6 addresses, removing the ipv6 related
> comments/docs section
> 
> Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>
> ---
>  block/gluster.c | 20 ++++++++++++--------
>  1 file changed, 12 insertions(+), 8 deletions(-)
> 
> diff --git a/block/gluster.c b/block/gluster.c
> index 40ee852..8a54ad4 100644
> --- a/block/gluster.c
> +++ b/block/gluster.c
> @@ -12,6 +12,7 @@
>  #include "block/block_int.h"
>  #include "qapi/error.h"
>  #include "qemu/uri.h"
> +#include "qemu/error-report.h"
>  
>  #define GLUSTER_OPT_FILENAME        "filename"
>  #define GLUSTER_OPT_DEBUG           "debug"
> @@ -134,12 +135,10 @@ static int parse_volume_options(GlusterConf *gconf, char *path)
>   *
>   * 'transport' specifies the transport type used to connect to gluster
>   * management daemon (glusterd). Valid transport types are
> - * tcp, unix and rdma. If a transport type isn't specified, then tcp
> - * type is assumed.
> + * tcp, unix. If a transport type isn't specified, then tcp type is assumed.

s/tcp, unix/tcp or unix/

>   *
>   * 'host' specifies the host where the volume file specification for
> - * the given volume resides. This can be either hostname, ipv4 address
> - * or ipv6 address. ipv6 address needs to be within square brackets [ ].
> + * the given volume resides. This can be either hostname, ipv4 address.

s/either hostname, ipv4 address/either a hostname or an ipv4 address/


> @@ -1048,6 +1046,12 @@ static BlockDriver bdrv_gluster_unix = {
>      .create_opts                  = &qemu_gluster_create_opts,
>  };
>  
> +/* rdma is deprecated (actually never supported for volfile fetch)
> + * lets maintain for the protocol compatibility, to make sure things

s/fetch) lets maintain for/fetch). Let's maintain it for/

> + * won't break immediately for now gluster+rdma will fall back to gluster+tcp

s/immediately for now gluster+rdma/immediately. For now, gluster+rdma/

> + * protocol with Warning

s/Warning/a warning/

> + * TODO: remove gluster+rdma interface support
> + */
>  static BlockDriver bdrv_gluster_rdma = {
>      .format_name                  = "gluster",
>      .protocol_name                = "gluster+rdma",
> 

Reviewed-by: Eric Blake <eblake@redhat.com>
Jeff Cody July 19, 2016, 6:46 p.m. UTC | #3
On Tue, Jul 19, 2016 at 10:27:31PM +0530, Prasanna Kumar Kalever wrote:
> gluster volfile server fetch happens through unix and/or tcp, it doesn't
> support volfile fetch over rdma, the rdma code may actually mislead,
> to make sure things do not break, for now we fallback to tcp when requested
> for rdma with a warning.
> 
> If you are wondering how this worked all these days, its the gluster libgfapi
> code which handles anything other than unix transport as socket/tcp, sad but
> true.
> 
> Also gluster doesn't support ipv6 addresses, removing the ipv6 related
> comments/docs section
> 
> Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>
> ---
>  block/gluster.c | 20 ++++++++++++--------
>  1 file changed, 12 insertions(+), 8 deletions(-)
> 
> diff --git a/block/gluster.c b/block/gluster.c
> index 40ee852..8a54ad4 100644
> --- a/block/gluster.c
> +++ b/block/gluster.c
> @@ -12,6 +12,7 @@
>  #include "block/block_int.h"
>  #include "qapi/error.h"
>  #include "qemu/uri.h"
> +#include "qemu/error-report.h"
>  
>  #define GLUSTER_OPT_FILENAME        "filename"
>  #define GLUSTER_OPT_DEBUG           "debug"
> @@ -134,12 +135,10 @@ static int parse_volume_options(GlusterConf *gconf, char *path)
>   *
>   * 'transport' specifies the transport type used to connect to gluster
>   * management daemon (glusterd). Valid transport types are
> - * tcp, unix and rdma. If a transport type isn't specified, then tcp
> - * type is assumed.
> + * tcp, unix. If a transport type isn't specified, then tcp type is assumed.
>   *
>   * 'host' specifies the host where the volume file specification for
> - * the given volume resides. This can be either hostname, ipv4 address
> - * or ipv6 address. ipv6 address needs to be within square brackets [ ].
> + * the given volume resides. This can be either hostname, ipv4 address.
>   * If transport type is 'unix', then 'host' field should not be specified.
>   * The 'socket' field needs to be populated with the path to unix domain
>   * socket.
> @@ -158,11 +157,8 @@ static int parse_volume_options(GlusterConf *gconf, char *path)
>   * file=gluster://1.2.3.4/testvol/a.img
>   * file=gluster+tcp://1.2.3.4/testvol/a.img
>   * file=gluster+tcp://1.2.3.4:24007/testvol/dir/a.img
> - * file=gluster+tcp://[1:2:3:4:5:6:7:8]/testvol/dir/a.img
> - * file=gluster+tcp://[1:2:3:4:5:6:7:8]:24007/testvol/dir/a.img
>   * file=gluster+tcp://host.domain.com:24007/testvol/dir/a.img
>   * file=gluster+unix:///testvol/dir/a.img?socket=/tmp/glusterd.socket
> - * file=gluster+rdma://1.2.3.4:24007/testvol/a.img
>   */
>  static int qemu_gluster_parseuri(GlusterConf *gconf, const char *filename)
>  {
> @@ -185,7 +181,9 @@ static int qemu_gluster_parseuri(GlusterConf *gconf, const char *filename)
>          gconf->transport = g_strdup("unix");
>          is_unix = true;
>      } else if (!strcmp(uri->scheme, "gluster+rdma")) {
> -        gconf->transport = g_strdup("rdma");
> +        gconf->transport = g_strdup("tcp");
> +        error_report("Warning: rdma feature is not supported falling "
> +                     "back to tcp");
>      } else {
>          ret = -EINVAL;
>          goto out;
> @@ -1048,6 +1046,12 @@ static BlockDriver bdrv_gluster_unix = {
>      .create_opts                  = &qemu_gluster_create_opts,
>  };
>  
> +/* rdma is deprecated (actually never supported for volfile fetch)
> + * lets maintain for the protocol compatibility, to make sure things
> + * won't break immediately for now gluster+rdma will fall back to gluster+tcp
> + * protocol with Warning
> + * TODO: remove gluster+rdma interface support
> + */
>  static BlockDriver bdrv_gluster_rdma = {
>      .format_name                  = "gluster",
>      .protocol_name                = "gluster+rdma",
> -- 
> 2.7.4
> 
>

Modulo Markus's and Eric's comments:

Reviewed-by: Jeff Cody <jcody@redhat.com>

(I'll make the comment changes on commit)
diff mbox

Patch

diff --git a/block/gluster.c b/block/gluster.c
index 40ee852..8a54ad4 100644
--- a/block/gluster.c
+++ b/block/gluster.c
@@ -12,6 +12,7 @@ 
 #include "block/block_int.h"
 #include "qapi/error.h"
 #include "qemu/uri.h"
+#include "qemu/error-report.h"
 
 #define GLUSTER_OPT_FILENAME        "filename"
 #define GLUSTER_OPT_DEBUG           "debug"
@@ -134,12 +135,10 @@  static int parse_volume_options(GlusterConf *gconf, char *path)
  *
  * 'transport' specifies the transport type used to connect to gluster
  * management daemon (glusterd). Valid transport types are
- * tcp, unix and rdma. If a transport type isn't specified, then tcp
- * type is assumed.
+ * tcp, unix. If a transport type isn't specified, then tcp type is assumed.
  *
  * 'host' specifies the host where the volume file specification for
- * the given volume resides. This can be either hostname, ipv4 address
- * or ipv6 address. ipv6 address needs to be within square brackets [ ].
+ * the given volume resides. This can be either hostname, ipv4 address.
  * If transport type is 'unix', then 'host' field should not be specified.
  * The 'socket' field needs to be populated with the path to unix domain
  * socket.
@@ -158,11 +157,8 @@  static int parse_volume_options(GlusterConf *gconf, char *path)
  * file=gluster://1.2.3.4/testvol/a.img
  * file=gluster+tcp://1.2.3.4/testvol/a.img
  * file=gluster+tcp://1.2.3.4:24007/testvol/dir/a.img
- * file=gluster+tcp://[1:2:3:4:5:6:7:8]/testvol/dir/a.img
- * file=gluster+tcp://[1:2:3:4:5:6:7:8]:24007/testvol/dir/a.img
  * file=gluster+tcp://host.domain.com:24007/testvol/dir/a.img
  * file=gluster+unix:///testvol/dir/a.img?socket=/tmp/glusterd.socket
- * file=gluster+rdma://1.2.3.4:24007/testvol/a.img
  */
 static int qemu_gluster_parseuri(GlusterConf *gconf, const char *filename)
 {
@@ -185,7 +181,9 @@  static int qemu_gluster_parseuri(GlusterConf *gconf, const char *filename)
         gconf->transport = g_strdup("unix");
         is_unix = true;
     } else if (!strcmp(uri->scheme, "gluster+rdma")) {
-        gconf->transport = g_strdup("rdma");
+        gconf->transport = g_strdup("tcp");
+        error_report("Warning: rdma feature is not supported falling "
+                     "back to tcp");
     } else {
         ret = -EINVAL;
         goto out;
@@ -1048,6 +1046,12 @@  static BlockDriver bdrv_gluster_unix = {
     .create_opts                  = &qemu_gluster_create_opts,
 };
 
+/* rdma is deprecated (actually never supported for volfile fetch)
+ * lets maintain for the protocol compatibility, to make sure things
+ * won't break immediately for now gluster+rdma will fall back to gluster+tcp
+ * protocol with Warning
+ * TODO: remove gluster+rdma interface support
+ */
 static BlockDriver bdrv_gluster_rdma = {
     .format_name                  = "gluster",
     .protocol_name                = "gluster+rdma",