Results 1 to 7 of 7
  1. #1
    Join Date
    May 2004
    Posts
    7

    Unanswered: DB2 TSA clusterNodeName restrictions?

    Hi,
    is anyone out there aware of any TSA clusterNodeName naming restrictions (or can point me to the right documentation please)?

    My specific problem is that I have a local host name with a dash inside it ("srv-d01"), and this seems to confuse TSA for some reason.
    When I run the db2haicu command, I get this error:

    FUNCTION: DB2 UDB, high avail services, handleEndElement, probe:4513
    MESSAGE : ECF=0x90000555=-1879046827=ECF_SQLHA_UPDATE_ATTR_FAILED
    Attribute update failed
    DATA #1 : String, 58 bytes
    Error during Domain creation from HA XML input processing.
    DATA #2 : String, 10 bytes
    hadrdomain
    DATA #3 : unsigned integer, 4 bytes
    2
    DATA #4 : String, 9 bytes
    srv-d01

    If I try to use an alias instead of the local host name, I get a different error:

    "The local hostname 'srv-d01' is not part of the host list passed in. Re-run the command and provide the local hostname in the host list."

    Any ideas?

    Regards,
    Damir

  2. #2
    Join Date
    Jul 2016
    Location
    Moscow
    Posts
    294
    Provided Answers: 45
    Hi,

    Try to use everywhere either short of full host names (and make the commands below to return them):

    hostname
    uname -n
    lsrpnode (if you have already created the cluster manually)
    HADR_LOCAL_HOST and HADR_REMOTE_HOST db2 database parameters
    your XML input file
    Regards,
    Mark.

  3. #3
    Join Date
    May 2004
    Posts
    7
    Thanks Mark!
    I've already sorted the server names (got rid of the dash in the server name(s)), but now I have another problem:

    2017-03-23-12.00.59.876303+000 E7466E938 LEVEL: Error
    PID : 11252 TID : 140377417869088 PROC : db2haicu
    INSTANCE: db2inst1 NODE : 000
    HOSTNAME: sir02d02
    FUNCTION: DB2 UDB, high avail services, sqlhaUpdateResource2, probe:450
    MESSAGE : ECF=0x90000555=-1879046827=ECF_SQLHA_UPDATE_ATTR_FAILED
    Attribute update failed
    DATA #1 : String, 35 bytes
    Error during vendor call invocation
    DATA #2 : SQLHA Cluster Session Handle, PD_TYPE_SQLHA_CLUSTER_HANDLE, 4120 bytes
    sqlhaClusterHandle->clusterHandle: 1
    sqlhaClusterHandle->clusterFlags: 0
    sqlhaClusterHandle->clusterErrorNum: 0
    sqlhaClusterHandle->errorMessage: Line # : 16497---Error Number: 33---2632-072 The operation cannot be performed because a majority of quorum nodes or configuration daemons is not currently active in the domain, or because the quorum of the domain is not currently satisfied.
    sqlhaClusterHandle->clusterCommand:

    That's the error message I get in db2diag.log when I run the "db2haicu -f [XML config file]" command...
    The cluster/node status is as follows:
    > lsrpdomain
    Name OpState RSCTActiveVersion MixedVersions TSPort GSPort
    db2HAdomain Online 3.1.5.5 No 12347 12348
    > lsrpnode -Q
    Name OpState RSCTVersion Quorum
    sir02d02 Online 3.1.5.5 Yes
    sir02d01 Offline 3.1.5.5 Yes

    So it seems that for some reason not all nodes in the cluster are activated during the setup?!?
    I'm checking the ports on the firewall (which are all open, according to the networking team)...

    Any thoughts on this one?

    Regards,
    Damir

  4. #4
    Join Date
    Jul 2016
    Location
    Moscow
    Posts
    294
    Provided Answers: 45
    What's the output of these commands from both nodes from root:

    lssrc -s IBM.ConfigRM
    lsrpnode
    samversion
    Regards,
    Mark.

  5. #5
    Join Date
    May 2004
    Posts
    7
    Hi Mark,
    I managed to solve the problem by opening up *all* ports between the two nodes, for all types of traffic (TCP, UDP)... not the best possible solution out there, but the only one that worked so far.
    The trouble is I am not sure now which ports are required for the normal TSA operation (and which ones I can safely shut down), as the documentation is very terse on that subject.
    I stumbled upon this article:
    https://developer.ibm.com/answers/qu...prsct-env.html
    and at first had the ports 657, 12347 and 12348 opened up (TCP, UDP) between the nodes, but that was apparently not sufficient.
    Are the ports 2001 and 2002 also absolutely necessary (or any other ones, not listed in the above link)?

    For the sake of completeness, here's the command output you asked for (it is the same on both nodes, except of course for the PID):

    $ lssrc -s IBM.ConfigRM
    Subsystem Group PID Status
    IBM.ConfigRM rsct_rm 25214 active

    $ lsrpnode
    Name OpState RSCTVersion
    sir02d02 Online 3.1.5.5
    sir02d01 Online 3.1.5.5

    $ samversion
    rsa322s008a 3.2.2.8 Jul 18 2014 10:02:26

    Thanks for your help!
    Regards,
    Damir

  6. #6
    Join Date
    May 2017
    Posts
    1
    I have no idea of it.

  7. #7
    Join Date
    Jul 2016
    Location
    Moscow
    Posts
    294
    Provided Answers: 45

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •