于 2011年09月22日 16:41, Eli 写道:

hi Osier:
于 2011年09月22日 15:08, Osier Yang 写道:

于 2011年09月22日 14:31, Eli 写道:

hi Osier :
于 2011年09月21日 17:07, Osier Yang 写道:

* src/storage/storage_backend_logical.c:

If a logical vol is created with multiple stripes. (e.g. --stripes 3),
the "device" field of lvs output will have multiple fileds which are
seperated by comma. It means the RE we write in the codes will not
work well anymore. E.g. (lvs output for a stripped vol, uses "#" as
seperator here):

test_stripes##fSLSZH-zAS2-yAIb-n4mV-Al9u-HA3V-oo9K1B#\
/dev/sdc1(10240),/dev/sdd1(0)#42949672960#4194304

The RE we uses:

     const char *regexes[] = {
         "^\\s*(\\S+),(\\S*),(\\S+),(\\S+)\\((\\S+)\\),(\\S+),([0-9]+),?\\s*$"
     };

This patch changes the seperator into "#" to fix the problem.

Related RHBZ: https://bugzilla.redhat.com/show_bug.cgi?id=727474
---
  src/storage/storage_backend_logical.c |    7 ++++---
  1 files changed, 4 insertions(+), 3 deletions(-)

diff --git a/src/storage/storage_backend_logical.c b/src/storage/storage_backend_logical.c
index 4f42047..45f77ad 100644
--- a/src/storage/storage_backend_logical.c
+++ b/src/storage/storage_backend_logical.c
@@ -187,19 +187,20 @@ virStorageBackendLogicalFindLVs(virStoragePoolObjPtr pool,
       *
       * NB can be multiple rows per volume if they have many extents
       *
-     * NB lvs from some distros (e.g. SLES10 SP2) outputs trailing "," on each line
+     * NB lvs from some distros (e.g. SLES10 SP2) outputs trailing
+     * @separator on each line
       *
       * NB Encrypted logical volumes can print ':' in their name, so it is
       *    not a suitable separator (rhbz 470693).
       */
      const char *regexes[] = {
-        "^\\s*(\\S+),(\\S*),(\\S+),(\\S+)\\((\\S+)\\),(\\S+),([0-9]+),?\\s*$"
+        "^\\s*(\\S+)#(\\S*)#(\\S+)#(\\S+)\\((\\S+)\\)#(\\S+)#([0-9]+)#?\\s*$"
      };
      int vars[] = {
          7
      };
      const char *prog[] = {
-        LVS, "--separator", ",", "--noheadings", "--units", "b",
+        LVS, "--separator", "#", "--noheadings", "--units", "b",
          "--unbuffered", "--nosuffix", "--options",
          "lv_name,origin,uuid,devices,seg_size,vg_extent_size",
          pool->def->source.name, NULL

I reproduced the bug :

[root@localhost bin]# ./virsh -d 5 pool-create-as vg_ssd logical --target /dev/vg_ssd
error: Failed to create pool vg_ssd
error: cannot open volume '/dev/vg_ssd/test_stripes,': No such file or directory

and then I tested this patch , it seems work well.

[root@localhost bin]# ./virsh -d 5 pool-create-as vg_ssd logical --target /dev/vg_ssd
Pool vg_ssd created

[root@localhost bin]# ./virsh pool-info vg_ssd
Name: vg_ssd
UUID: c45cc84e-7879-cc15-ee78-2d2dda6b531d
State: running
Persistent: no
Autostart: no
Capacity: 200.00 MB
Allocation: 152.00 MB
Available: 48.00 MB

Thanks for the testing, Eli,

Could you also check what's the vol XML?  I want to confirm if the
"<extents>" in "<source><device></device></source>"displays well,
though it looks to me there is no "<path>" element defined for "<extents>"
in the storage vol schema yet.

1. my vol XML is like this:

  virsh # vol-dumpxml test_stripes --pool vg_ssd
<volume>
  <name>test_stripes</name>
  <key>1pc6Gf-1hn2-WGnw-ASKt-uX6w-xUed-qERq50</key>
  <source>
    <device path='/dev/sdb(0),/dev/sdc'>

This is expected, and is what's Daniel Berrange worried about.
 

      <extent start='0' end='159383552'/>
    </device>
  </source>

Thanks
Osier

-- 
best regards
eli