Not able to evacuate disk

[YC]

Hi,

We are using Veritas Storage Foundation Suite V4.1 in Solaris 10 supporting Oracle DB. Currently, one of the file system (see below) /u03 is supported by vol04 vol that has 2 disks as follows:

:

sd u03-01       vol04-01     u03      0        20971520 0         TagmaStore-USP0_3 ENA
sd u03-02-01    vol04-01     u03-02   0        60817408 20971520  TagmaStore-USP0_18 ENA
 

# df -h /u03
Filesystem             size   used  avail capacity  Mounted on
/dev/vx/dsk/oradg/vol04
                        39G   5.6G    31G    16%    /u03
 

sd u03-01 is 10GB and sd u-3-02-01 is 40GB. Now the usage for /u03 is 5.6GB and I would like to evacuate u03-01 out and move the data to u03-02-01 but got the message:

VxVM vxassist ERROR V-5-1-438 Cannot allocate space to replace subdisks

Any advice is appreciated and thanks in advance.

[Marianne]

What is the volume layout? If striped, you will need another disk to move data off u03-01.

Please post output of 'vxprint -ht -g <diskgroup> vol04'

PS: Are you aware of the fact that SF 4.1 reached EOSL last year?
See http://www.symantec.com/business/support/index?page=releasedetails&key=15107

and https://www-secure.symantec.com/connect/forums/heads-technical-support-storage-foundation-4x-ends-july-31-2011-forum-copy

[YC]

Hi,

# vxprint -ht -g oradg vol04
V  NAME         RVG/VSET/CO  KSTATE   STATE    LENGTH   READPOL   PREFPLEX UTYPE
PL NAME         VOLUME       KSTATE   STATE    LENGTH   LAYOUT    NCOL/WID MODE
SD NAME         PLEX         DISK     DISKOFFS LENGTH   [COL/]OFF DEVICE   MODE
SV NAME         PLEX         VOLNAME  NVOLLAYR LENGTH   [COL/]OFF AM/NM    MODE
SC NAME         PLEX         CACHE    DISKOFFS LENGTH   [COL/]OFF DEVICE   MODE
DC NAME         PARENTVOL    LOGVOL
SP NAME         SNAPVOL      DCO
EX NAME         ASSOC        VC                       PERMS    MODE     STATE
SR NAME         KSTATE

v  vol04        -            ENABLED  ACTIVE   81788928 SELECT    -        fsgen
pl vol04-01     vol04        ENABLED  ACTIVE   81788928 CONCAT    -        RW
sd u03-01       vol04-01     u03      0        20971520 0         TagmaStore-USP0_3 ENA
sd u03-02-01    vol04-01     u03-02   0        60817408 20971520  TagmaStore-USP0_18 ENA

[Marianne]

OK, you have a concatenated disk.

Please show free space on each disk in dg:

vxdg -g oradg free

[YC]

Hi,

# vxdg -g oradg free
DISK         DEVICE       TAG          OFFSET    LENGTH    FLAGS
exl-1        TagmaStore-USP0_22 TagmaStore-USP0_22 314490880 768       -
exl-2        TagmaStore-USP0_23 TagmaStore-USP0_23 20883456  1792      -
exlibris     TagmaStore-USP0_7 TagmaStore-USP0_7 20971520  121088    -
exlibris-02  TagmaStore-USP0_21 TagmaStore-USP0_21 144703488 2010880   -
exp          TagmaStore-USP0_11 TagmaStore-USP0_11 41859072  256       -
rman         TagmaStore-USP0_2 TagmaStore-USP0_2 6422528   1280      -
u01          TagmaStore-USP0_4 TagmaStore-USP0_4 10485760  123648    -
u02          TagmaStore-USP0_5 TagmaStore-USP0_5 115343360 121088    -
u02-02       TagmaStore-USP0_12 TagmaStore-USP0_12 20883456  1792      -
u02-03       TagmaStore-USP0_13 TagmaStore-USP0_13 20761792  123456    -
u02-04       TagmaStore-USP0_14 TagmaStore-USP0_14 20552096  333152    -
u02-05       TagmaStore-USP0_16 TagmaStore-USP0_16 83798016  1792      -
u02-06       TagmaStore-USP0_17 TagmaStore-USP0_17 207618032 2010896   -
u03          TagmaStore-USP0_3 TagmaStore-USP0_3 20971520  121088    -
u03-02       TagmaStore-USP0_18 TagmaStore-USP0_18 81788768  2011040   -
u04          TagmaStore-USP0_6 TagmaStore-USP0_6 104857600 131328    -
u04-02       TagmaStore-USP0_19 TagmaStore-USP0_19 10276048  126000    -
u05          TagmaStore-USP0_1 TagmaStore-USP0_1 20971520  121088    -
u05-02       TagmaStore-USP0_20 TagmaStore-USP0_20 20761808  123440    -
u06          TagmaStore-USP0_0 TagmaStore-USP0_0 52428800  121088    -
u07          TagmaStore-USP0_8 TagmaStore-USP0_8 20845696  39552     -
u07-02       TagmaStore-USP0_15 TagmaStore-USP0_15 20883456  1792      -
u08          TagmaStore-USP0_9 TagmaStore-USP0_9 20845696  39552     -
u09          TagmaStore-USP0_10 TagmaStore-USP0_10 20845696  39552     -
 

[Marianne]

u03          TagmaStore-USP0_3 TagmaStore-USP0_3 20971520  121088    -

u03-02       TagmaStore-USP0_18 TagmaStore-USP0_18 81788768  2011040   -

Space occupied on TagmaStore-USP0_18:  81788768

Free space on TagmaStore-USP0_18: 2011040

Space occupied on TagmaStore-USP0_3: 20971520

20971520  (needed)
2011040    (free)

So, you can see that that TagmaStore-USP0_18 does not have enough free space....

[YC]

Hi,

Why it shows the space is sufficient from the output of df (see below, avail is 31GB) but insufficient from the output of vgdg -g <dg> free?   

# df -h /u03

Filesystem             size   used  avail capacity  Mounted on
/dev/vx/dsk/oradg/vol04
                        39G   5.6G    31G    16%    /u03

[Gaurav_S · Accepted solution]

I believe what you are trying to achieve is incorrect ... reason being ..

you have a concat volume & certainly the data would lie on first plex or first sd i.e u03-01 ... now you are trying to run an evacuate on the first disk itself ..... that means veritas would try to find another disk of having 10G free space (& I assume here it won't pick u03-02 as it is already in use on second plex, moreover it doesn't have a free space (only 0.95G left)

Point to note, to make an evacuation, you need exact 10G of space available & if you notice the last coulmn of all the disks in vxdg free output .... you don't have any disk with that much free space ....

u03          TagmaStore-USP0_3 TagmaStore-USP0_3 20971520       121088 -----> (0.057G)
u03-02       TagmaStore-USP0_18 TagmaStore-USP0_18 81788768   2011040---->  (0.95G)

Solution what would work for you is,

1. Shrink the volume (vxresize) to 8G or 9G.... vxvm may attempt to free up your second subdisk here i.e u03-02 (make sure you have all backups in place to be on safe side)

2. Once above disk is free, mirror the original volume from u03-02, so a new subdisk from u03-02 will be created & a new plex as well & mirroed to u03-01

then you can move out to the old plex or new plex which you want  ...  vxevac won't be a good choice here since you want to move the data within subdisks of a single plex ...

Hope this answers ...

Gaurav

[Marianne]

In addition to my and Gaurav's posts, please understand that 'df' output shows you filesystem usage, not volume/disk usage.

As per Gaurav's post, you can shrink the filesystem provided this is a vxfs filesystem, and then proceed with rest of Gaurav's advice.

[YC]

Hi,

Would you like to share how to convert 121088 -----> (0.057G)?

[mikebounds]

Divide by 2048 to get MB:
121088 / 2048 = 59.125 MB

Divide again by 1024 to get GB:

121088 / (2048 x 1024) = 0.577GB

Mike

[mikebounds]

A few questions:

1. What are you trying to achieve over all - are you trying to free up space in diskgroup so you can extend other volumes in the diskgroup, or free up space so you can remove from diskgroup (to give to another diskgroup or give back space to array).
    
2. Do you need to free up disk u03-01 in particular or can you free any disk in the diskgroup.
    
3. Do you specifically want to shrink /u03 filesystem from 40GB to 30GB or would another size be more useful?
    
4. Do you layout volumes on particular LUNS or do you let Volume Manager put them where it wants to?
    
5. Do you use thin LUNs?

Would be useful to see full vxprint output.

Mike