Proposed change to vstart.sh

Commit Message

Joe Buck Oct. 21, 2012, 9:31 p.m. UTC

It looks like vstart.sh does not work without authx enabled. Given that, 
I'd propose to change the default to having cephx be enabled and then 
use the -x flag to disable cephx.

Interestingly enough, the help output for vstart.sh already states that 
the -x flag disables authx (which is not the case at present, but is 
after this patch is applied).

I've attached a patch to vstart that 1) changes the mention of authx to 
cephx for consistency and 2) defaults vstart.sh to using cephx with the 
-x flag now serving to disable cephx.

Best,
-Joe Buck

Comments

Sage Weil Oct. 21, 2012, 9:35 p.m. UTC | #1

On Sun, 21 Oct 2012, Joe Buck wrote:
> It looks like vstart.sh does not work without authx enabled. Given that, I'd
> propose to change the default to having cephx be enabled and then use the -x
> flag to disable cephx.
> 
> Interestingly enough, the help output for vstart.sh already states that the -x
> flag disables authx (which is not the case at present, but is after this patch
> is applied).
> 
> I've attached a patch to vstart that 1) changes the mention of authx to cephx
> for consistency and 2) defaults vstart.sh to using cephx with the -x flag now
> serving to disable cephx.

Sounds like a good change, but let's make -x do nothing and -X disable 
cephx, so that existing habitual typing of -x doesn't lead to failure.

Also, we should fix the auth none usage at some point.  :)

sage
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to [e-mail]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Dan Mick Oct. 22, 2012, 6:05 a.m. UTC | #2

On 10/21/2012 02:35 PM, Sage Weil wrote:
> On Sun, 21 Oct 2012, Joe Buck wrote:
>> It looks like vstart.sh does not work without authx enabled. Given that, I'd
>> propose to change the default to having cephx be enabled and then use the -x
>> flag to disable cephx.
>>
>> Interestingly enough, the help output for vstart.sh already states that the -x
>> flag disables authx (which is not the case at present, but is after this patch
>> is applied).
>>
>> I've attached a patch to vstart that 1) changes the mention of authx to cephx
>> for consistency and 2) defaults vstart.sh to using cephx with the -x flag now
>> serving to disable cephx.
> Sounds like a good change, but let's make -x do nothing and -X disable
> cephx, so that existing habitual typing of -x doesn't lead to failure.
>
> Also, we should fix the auth none usage at some point.  :)
>

Surely it's not hard to fix the non-cephx case?  I though slang had done 
that, actually.

--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to [e-mail]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Patch

diff --git a/src/vstart.sh b/src/vstart.sh
index 7aaad49..b23b83a 100755
--- a/src/vstart.sh
+++ b/src/vstart.sh
@@ -21,7 +21,7 @@  ip=""
 nodaemon=0
 smallmds=0
 overwrite_conf=1
-cephx=0
+cephx=1 #turn cephx on by default
 
 MON_ADDR=""
 
@@ -43,7 +43,7 @@  usage=$usage"\t--nodaemon: use ceph-run as wrapper for mon/osd/mds\n"
 usage=$usage"\t--smallmds: limit mds cache size\n"
 usage=$usage"\t-m ip:port\t\tspecify monitor address\n"
 usage=$usage"\t-k keep old configuration files\n"
-usage=$usage"\t-x disable authx\n"
+usage=$usage"\t-x disable cephx\n"
 usage=$usage"\t-o config\t\t add extra config parameters to mds section\n"
 
 usage_exit() {
@@ -117,7 +117,7 @@  case $1 in
 	    shift
 	    ;;
     -x )
-	    cephx=1
+	    cephx=0
 	    ;;
     -k )
 	    overwrite_conf=0