You are here

Installation Veritas Volume Manager 4.0 mit Veritas Cluster 3.5

 

 

Installation VCS 4.0 mit Veritas Cluster 3.5

 

 

  1. Installation Solaris 8
  2. Installation Veritas Volume Manager 4.0, mit encapsulieren von rootdg
  3. Veritas Cluster 3.5 Packes über Factory verteilen und installieren
  4. .Zusatz Path eintag in profile eintragen:
    PATH=/usr/sbin:/sbin:/opt/VRTSvcs/bin

  5. EEPROM Eintrag:
    #
    eeprom local-mac-address?=true

     
  6. Hartbeat prüfen

    Sender:
    #
    /opt/VRTSllt/getmac /dev/qfe:0

    /dev/qfe:0 08:00:20:E7:DE:B2


    # /opt/VRTSllt/dlpiping -s /dev/qfe:0


    Empfaenger:
    #
    /opt/VRTSllt/dlpiping -c /dev/qfe:0 08:00:20:E7:DE:B2

    08:00:20:E7:DE:B2 is alive           ## war erfolgreich

    # /opt/VRTSllt/dlpiping -c /dev/hme:1 08:00:20:E7:DE:B2
    no response from 08:00:20:E7:DE:B2  
    ## bei Verbindungsprobleme

 

 

  1. - edit /.rhosts auf beiden Rechner
    - rsh in der /etc/inetd.conf freigeben und rsh testen (muss nicht umgedingt sein )

  2. Dateien /etc/llttab, /etc/llthosts, /etc/gabtab von einer anderen Rechner holen und Modifiezieren

    # /etc/llthosts
    0 DEFX0WF0
    1 DEFX0WF1

    #

    /etc/llttab

    set-node /etc/VRTSvcs/conf/sysname
    set-cluster 21
    link qfe1 /dev/qfe:1 - ether - -
    link qfe2 /dev/qfe:2 - ether - -
    link-lowpri hme0 /dev/hme:0 - ether - -
    exclude 3-31

    #

    more gabtab

    /sbin/gabconfig -cx -n2

  1. Minimale “main.cf” create:

    #

    vi /etc/VRTSvcs/conf/config/main.cf
    include "types.cf"

    cluster test_vcs_WM (

    UserNames = { admin = Z35jyNrF5YVu6 }
    Administrators = { admin }
    CounterInterval = 5
    )

    system DEFX0WF0 (

    )

    system DEFX0WF1 (

    )
    group dataSG (

    SystemList = { DEFX0WF0 = 1, DEFX0WF1 = 2 }
    AutoStart = 1
    Parallel = 0
    AutoStartList = { DEFX0WF0, DEFX0WF1 }
    )

  2. Datei sysname create und edit mit den eigenen Hostnamen

    #
    more /etc/VRTSvcs/conf/sysname

    DEFX0WF0

  1. Starten LLT Network Protocol auf beiden Rechner

    #
    /etc/rc2.d/S70llt start

    Starting LLT...
    Starting LLT done.

 

  1. Starten von GAB Broadcast Protocol auf beiden Rechner

    #

    /etc/rc2.d/S92gab start

    Starting GAB...
    Starting GAB done.

  1. Starten vom Cluster mit „ hastart „ auf beide Nodes

- Prüfen Log Datei /var/VRTSvcs/log/engine_A.log
-
hastatus –sum ; hastatus                                      ## Prüfen
 -  hagrp -autoenable dataSG -sys DEFX0WF0       ## Auf beiden Rechnern

  1. Anlegen von Disks Group, Volume und Filesysteme unter Veritas Volume Manager.

    # vxdisk list
           DEVICE TYPE DISK GROUP STATUS

c1t9d0s2 auto:none - - online invalid
c1t10d0s2 auto:none - - online invalid
c2t9d0s2 auto:none - - online invalid
c2t10d0s2 auto:none - - online invalid

# vxdisksetup -i c1t9d0
# vxdisksetup -i c1t10d0
# vxdisksetup -i c2t10d0
# vxdisksetup -i c2t9d0

# vxdisk list
     
DEVICE TYPE DISK GROUP STATUS

c1t9d0s2 auto:cdsdisk - - online
c1t10d0s2 auto:cdsdisk - - online
c2t9d0s2 auto:cdsdisk - - online
c2t10d0s2 auto:cdsdisk - - online 

# vxdg init dataDG data01=c1t9d0
# vxdg -g dataDG adddisk data02=c1t10d0
# vxdg -g dataDG adddisk data03=c2t9d0
# vxdg -g dataDG adddisk data04=c2t10d0

# vxdisk list
                  DEVICE TYPE DISK GROUP STATUS

c1t9d0s2 auto:cdsdisk data01 dataDG online
c1t10d0s2 auto:cdsdisk data02 dataDG online
c2t9d0s2 auto:cdsdisk data03 dataDG online
c2t10d0s2 auto:cdsdisk data04 dataDG online

 

-- VCS nimmt disksgroup unter Kontrolle indem er sie einmal deportiert (Option noautostart) und wieder importiert.

# vxdg deport dataDG

# vxdg -t import dataDG

# vxprint -l dataDG |grep autoimport

info: dgid=1132739091.22.DEFX0WF0 noautoimport

# vxassist -b -g dataDG make vol01 4g layout=mirror data01 data03

 

# vxprint -g dataDG

TY NAME ASSOC KSTATE LENGTH PLOFFS STATE TUTIL0 PUTIL0

dg dataDG dataDG - - - - - - 

dm data01 c1t9d0s2 - 35365968 - - - -
dm data02 c1t10d0s2 - 35365968 - - - -
dm data03 c2t9d0s2 - 35365968 - - - -
dm data04 c2t10d0s2 - 35365968 - - - -

 

v vol01 fsgen ENABLED 8388608 - SYNC - -

pl vol01-01 vol01 ENABLED 8388608 - ACTIVE - -
sd data01-01 vol01-01 ENABLED 8388608 0 - - -
pl vol01-02 vol01 ENABLED 8388608 - ACTIVE - -
sd data03-01 vol01-02 ENABLED 8388608 0 - - -

# mkdir /gogo

# mkfs -F vxfs /dev/vx/rdsk/dataDG/vol01

# fsck –y –F vxfs /dev/vx/rdsk/dataDG/vol01

# mount -F vxfs /dev/vx/dsk/dataDG/vol01 /gogo

  1. Script „auto_vcs_crete“ ausführen der die Ganzen Resource kreiert.
    Die folgende Meldung ist zu Ignorieren:

VCS:10245:Resource added
NameRule and Enabled attributes must be set before agent monitors

more /auto_vcs_create

#add Resource Group

##hagrp -add <Resource Group>

##hagrp -modify <Resource Group> SystemList <Nodename1> 0 <NodeName2> 1

##hagrp -modify <Resource Group> AutoStartList <Node to start on boot>

##hagrp -modify <Resource Group> AutoStart 1

haconf -makerw

#Device

hares -add NICres NIC dataSG

hares -modify NICres Device hme0

hares -modify NICres Critical 0

hares -modify NICres Enabled 1

 

#Interface IP

hares -add IPres IP dataSG

hares -modify IPres Device hme0

## Log Cluster Ip Adress

hares -modify IPres Address 130.10.148.83

hares -modify IPres NetMask 255.255.255.0

hares -modify IPres Critical 0

hares -modify IPres Enabled 1

 

#Disk Group

hares -add dataDGres DiskGroup dataSG

hares -modify dataDGres DiskGroup dataDG

hares -modify dataDGres Critical 1

hares -modify dataDGres StopVolumes 1

hares -modify dataDGres StartVolumes 1

hares -modify dataDGres Enabled 1

 

 

#Volume

hares -add Volres Volume dataSG

hares -modify Volres Critical 0

hares -modify Volres Volume vol01

hares -modify Volres DiskGroup dataDG

hares -modify Volres Enabled 1

 

#Mount points

hares -add Mountres Mount dataSG

hares -modify Mountres BlockDevice /dev/vx/dsk/dataDG/vol01

hares -modify Mountres MountPoint /gogo

hares -modify Mountres FSType vxfs

hares -modify Mountres FsckOpt %-y

hares -modify Mountres Critical 0

hares -modify Mountres Enabled 1

 

#Build dependencies

hares -link IPres NICres

hares -link dataDGres IPres

hares -link Volres dataDGres

hares -link Mountres Volres

 

haconf -dump –makero

 



  1. Zum Schluss die Service Group starten.
    # hagrp -online dataSG -sys DEFX0WF0


  2. Cluster Switch Test.
    #
    hagrp -switch dataSG -to DEFX0WF1