communications error when starting with multiple "nodes"

Thread view:

Enable EMail Alerts

Start New Thread

Thread rating:

Mairhtin O'Feannag - 25 Oct 2004 05:30 GMT

Hello,

We have two machines we wish to use DPF. They are both RH ES 2.1, with
DB2 8.2.

I read the documentation CAREFULLY, and added the following line to my
db2nodes.cfg file :

1 egret 0

So the file now looks like :

0 hawk 0
1 egret 0

Both machines hawk and egret are on the same network and have the
following IFCONFIGs :

Hawk :
eth0 Link encap:Ethernet HWaddr 00:B0:D0:49:52:07
inet addr:172.16.x.xx Bcast:172.16.x.255 Mask:255.255.252.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:977972 errors:0 dropped:0 overruns:0 frame:0
TX packets:79625 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:88047038 (83.9 Mb) TX bytes:17105677 (16.3 Mb)
Interrupt:26 Base address:0xd000

Egret:
eth0 Link encap:Ethernet HWaddr 00:B0:D0:20:94:F0
inet addr:172.16.x.xx Bcast:172.16.x.255 Mask:255.255.252.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:1414618 errors:0 dropped:0 overruns:0 frame:0
TX packets:97243 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:128729435 (122.7 Mb) TX bytes:7877213 (7.5 Mb)
Interrupt:16 Base address:0xd000

so that they are both on the 172.16 network.

When I do a DB2START on Hawk (the coordinator node) I get :

10/24/2004 21:22:34 0 0 SQL6048N A communication error occurred
during START or STO.
10/24/2004 21:22:35 1 0 SQL6048N A communication error occurred
during START or STO.
SQL1032N No start database manager command was issued. SQLSTATE=57019

And I am at a loss as to why. The documentation says this is the way to
point to the various machines. The /etc/services entries are identical
on both machines too.

What am I not doing? Or doing wrong?????

Here is a portion of the DB2DIAG.LOG file, the tail output from the
db2start command ....

2004-10-24-21.04.17.337570-420 I2131977G409 LEVEL: Error
PID : 27061 TID : 8192 PROC : db2start
INSTANCE: db2gen NODE : 000
FUNCTION: DB2 UDB, oper system services, sqloPdbExecuteRemoteCmd,
probe:50
DATA #1 : Hexdump, 21 bytes
0x436BE300 : 6C6F 6361 6C68 6F73 742E 6C6F 6361 6C64 localhost.locald
0x436BE310 : 6F6D 6169 6E omain

2004-10-24-21.04.17.338436-420 I2132387G407 LEVEL: Error
PID : 27061 TID : 8192 PROC : db2start
INSTANCE: db2gen NODE : 000
FUNCTION: DB2 UDB, oper system services, sqloPdbExecuteRemoteCmd,
probe:50
DATA #1 : Hexdump, 19 bytes
0xBFFE899C : 5065 726D 6973 7369 6F6E 2064 656E 6965 Permission denie
0xBFFE89AC : 642E 0A d..

2004-10-24-21.04.17.339018-420 I2132795G352 LEVEL: Severe
PID : 27061 TID : 8192 PROC : db2start
INSTANCE: db2gen NODE : 000
FUNCTION: DB2 UDB, oper system services, sqloPdbOpenConnection, probe:10
MESSAGE : DiagData
DATA #1 : Hexdump, 4 bytes
0xBFFE8E00 : 1200 0F81 ....

2004-10-24-21.04.17.346717-420 I2133148G260 LEVEL: Error
PID : 27061 TID : 8192 PROC : db2start
INSTANCE: db2gen NODE : 000
FUNCTION: DB2 UDB, oper system services, sqloPdbCloseSocket, probe:5
MESSAGE : errno = 9

2004-10-24-21.04.17.347170-420 I2133409G372 LEVEL: Error
PID : 27061 TID : 8192 PROC : db2start
INSTANCE: db2gen NODE : 000
FUNCTION: DB2 UDB, oper system services, sqloPdbFreeCmdHandle, probe:10
MESSAGE : Error closing socket, OSSrc =
DATA #1 : Hexdump, 4 bytes
0xBFFE8E9C : 2500 0F87 %...

2004-10-24-21.04.17.347555-420 I2133782G260 LEVEL: Error
PID : 27061 TID : 8192 PROC : db2start
INSTANCE: db2gen NODE : 000
FUNCTION: DB2 UDB, oper system services, sqloPdbCloseSocket, probe:5
MESSAGE : errno = 9

2004-10-24-21.04.17.348003-420 I2134043G372 LEVEL: Error
PID : 27061 TID : 8192 PROC : db2start
INSTANCE: db2gen NODE : 000
FUNCTION: DB2 UDB, oper system services, sqloPdbFreeCmdHandle, probe:15
MESSAGE : Error closing socket, OSSrc =
DATA #1 : Hexdump, 4 bytes
0xBFFE8E9C : 2500 0F87 %...

2004-10-24-21.04.17.769395-420 I2134416G346 LEVEL: Error
PID : 27061 TID : 8192 PROC : db2start
INSTANCE: db2gen NODE : 000
FUNCTION: DB2 UDB, oper system services, sqloPdbExecuteRemoteCmd,
probe:50
DATA #1 : Hexdump, 15 bytes
0x436BE300 : 6567 7265 742E 6567 6372 632E 6F72 67 egret.egcrc.org

2004-10-24-21.04.17.769819-420 I2134763G407 LEVEL: Error
PID : 27061 TID : 8192 PROC : db2start
INSTANCE: db2gen NODE : 000
FUNCTION: DB2 UDB, oper system services, sqloPdbExecuteRemoteCmd,
probe:50
DATA #1 : Hexdump, 19 bytes
0xBFFE899C : 5065 726D 6973 7369 6F6E 2064 656E 6965 Permission denie
0xBFFE89AC : 642E 0A d..

2004-10-24-21.04.17.770364-420 I2135171G352 LEVEL: Severe
PID : 27061 TID : 8192 PROC : db2start
INSTANCE: db2gen NODE : 000
FUNCTION: DB2 UDB, oper system services, sqloPdbOpenConnection, probe:10
MESSAGE : DiagData
DATA #1 : Hexdump, 4 bytes
0xBFFE8E00 : 1200 0F81 ....

2004-10-24-21.04.17.772366-420 I2135524G260 LEVEL: Error
PID : 27061 TID : 8192 PROC : db2start
INSTANCE: db2gen NODE : 000
FUNCTION: DB2 UDB, oper system services, sqloPdbCloseSocket, probe:5
MESSAGE : errno = 9

2004-10-24-21.04.17.772715-420 I2135785G372 LEVEL: Error
PID : 27061 TID : 8192 PROC : db2start
INSTANCE: db2gen NODE : 000
FUNCTION: DB2 UDB, oper system services, sqloPdbFreeCmdHandle, probe:10
MESSAGE : Error closing socket, OSSrc =
DATA #1 : Hexdump, 4 bytes
0xBFFE8E9C : 2500 0F87 %...

2004-10-24-21.04.17.773089-420 I2136158G260 LEVEL: Error
PID : 27061 TID : 8192 PROC : db2start
INSTANCE: db2gen NODE : 000
FUNCTION: DB2 UDB, oper system services, sqloPdbCloseSocket, probe:5
MESSAGE : errno = 9

2004-10-24-21.04.17.773432-420 I2136419G372 LEVEL: Error
PID : 27061 TID : 8192 PROC : db2start
INSTANCE: db2gen NODE : 000
FUNCTION: DB2 UDB, oper system services, sqloPdbFreeCmdHandle, probe:15
MESSAGE : Error closing socket, OSSrc =
DATA #1 : Hexdump, 4 bytes
0xBFFE8E9C : 2500 0F87 %...

Reply to this Message

Ian - 25 Oct 2004 14:48 GMT

> I read the documentation CAREFULLY, and added the following line to my
> db2nodes.cfg file :

What does your .rhosts look like? Have you enabled rsh
(/etc/xinetd.d/rshd)?

You should be able to successfully execute the command 'rah date' and
get the date from each server before you try to start DB2.

Reply to this Message

mairhtin o'feannag - 25 Oct 2004 18:16 GMT

Well,

If .rhosts is a needed file, it isn't in the doc. :(

I don't seem to *HAVE* an .rhosts file on either system. Should they be
the same on both boxes?

M

>> I read the documentation CAREFULLY, and added the following line to
>> my db2nodes.cfg file :
[quoted text clipped - 9 lines]
> World! >100,000 Newsgroups ---= East/West-Coast Server Farms - Total
> Privacy via Encryption =---

Reply to this Message

Ian - 25 Oct 2004 18:54 GMT

> Well,
>
> If .rhosts is a needed file, it isn't in the doc. :(
>
> I don't seem to *HAVE* an .rhosts file on either system. Should they be
> the same on both boxes?

The instance owner home directory needs to be NFS mounted across all
servers. So, yes, the .rhosts should be the same.

Have you read the "Quick Beginnings for DB2 Servers" guide? This stuff
is all documented pretty clearly.

Reply to this Message

Mairhtin O'Feannag - 26 Oct 2004 16:31 GMT

Well.. folks,... you're gonna LOVE this one.

It turns out that KSH is *REQUIRED* when doing DPF. This is because IBM
still have scripts that run with

#!/bin/ksh

as the first line, and if you have no KSH, the script fails with a "bad
interpreter" error.

Then DB2 throws the infamous TCP/IP Communications error at a DB2START.

VERY VERY informative message.

I feel like I'm a detective. Even level 2 of IBM support missed this
one.

Time for IBM to use better messages, I'd say. Passing the "bad
interpreter" message along to the right scripts and then to the user
would have saved DAYS of work.

But that's just my opinion.

Mairhtin

> Hello,
>
[quoted text clipped - 159 lines]
> DATA #1 : Hexdump, 4 bytes
> 0xBFFE8E9C : 2500 0F87 %...

Reply to this Message

Ian - 26 Oct 2004 18:00 GMT

> Well.. folks,... you're gonna LOVE this one.
>
> It turns out that KSH is *REQUIRED* when doing DPF. This is because IBM
> still have scripts that run with

If you read the Installation Requirements you would know this. It's
no big secret.

http://publib.boulder.ibm.com/infocenter/db2help/topic/com.ibm.db2.udb.doc/start
/r0008865.htm

With the complexity from DPF, it's imperative that you RTFM. As I said
yesterday, the Quick Beginnings guide documents all of this.

Reply to this Message

mairhtin o'feannag - 27 Oct 2004 22:28 GMT

Ian,

While I appreciate your help, much to YOUR chagrin, it is NOT in the
quick beginnings book. I just read EVERY BIT of the linux sections.
EVERY WORD. I am reading from Quick Beginnings for DB2 Servers,
publication GC09-4836-00

NOT ONE WORD about ksh being required.

Sorry to disappoint, but I am rather thorough when reading.

When I was on the phone with IBM tech support, I mentioned that there
seemed in the past to be a requirement that pdksh be installed, and I was
told BY THEM "that's only for the install, we don't use ksh for our
scripts anymore". That's from one Rafat Odeh at IBM. IBM.

And after reading the db2diag.log file entry, the Level 2 "engine"
engineer concurred and flatly stated that ksh is NOT NEEDED.

So what's a poor girl to think, eh?

M

>> Well.. folks,... you're gonna LOVE this one.
>>
[quoted text clipped - 14 lines]
> World! >100,000 Newsgroups ---= East/West-Coast Server Farms - Total
> Privacy via Encryption =---

Reply to this Message

Jan M. Nelken - 27 Oct 2004 22:55 GMT

> Ian,
>
[quoted text clipped - 6 lines]
>
> Sorry to disappoint, but I am rather thorough when reading.

Did you read Page 69 of Quick Beginning for DB2 Servers (GC09-4836-01)?

Jan M. Nelken

Reply to this Message

Jan M. Nelken - 27 Oct 2004 22:58 GMT

> Ian,
>
[quoted text clipped - 37 lines]
>>World! >100,000 Newsgroups ---= East/West-Coast Server Farms - Total
>>Privacy via Encryption =---

Reply to this Message

Jan M. Nelken - 27 Oct 2004 23:06 GMT

> Ian,
>
> While I appreciate your help, much to YOUR chagrin, it is NOT in the
> quick beginnings book. I just read EVERY BIT of the linux sections.
> EVERY WORD. I am reading from Quick Beginnings for DB2 Servers,
> publication GC09-4836-00

Sorry for interrupted messages - I press Send too soon...

From your original messsage:

> Hello,
>
> We have two machines we wish to use DPF. They are both RH ES 2.1, with
> DB2 8.2.

So - you were installing Db2 V8.2 - and you were reading QB for DB2 Servers for
V8.1!

Try again - with Quick Beginnings for DB2 Servers (GC09-4836-01) for V8.2.
Specifically - read carefully page 69 of that publication.

You can find DB2 UDB English manuals here:

http://www-306.ibm.com/software/data/db2/udb/support/manualsv8.html

There are *two* columns for each of the items; one column has V8.1 title, one
has V8.2 title.
Click on PDF symbol in *column with V8.2 title* for item "Qucik beginnings for
DB2 Servers".

Jan M. Nelken

Reply to this Message