ODBC listener gone off to never-never land?

hp

ODBC listener gone off to never-never land?

Postby Emerson, Tom » Fri, 17 Sep 2004 08:16:36 GMT

We're seeing something unusual "from time to time" with the ODBC listener -- in particular, here is a glance screenshot of the main PIN:

HPB1787 B.10.07 HP GlancePlus/iX  16:06:20       Process (Pin 176 - #J7923)
                                 + 0: 0:29                     current avg high 
CPU   SB        *   B..........................................|  27%  18%  36%
Disc  MS               *               SIB         B...........|   79   30   79
Mem   M MS      SD     DF                      F.*.............|  74%  75%  75%
                         Process Selection Summary                              
CPU   *..............................<  0.0% |  NMPRG ODBCLNSE.ODBCSE.SYS
Disc  *..............................<  0.0  |  Elapsed  |  CPU-Tot  |  Disc-Tot
WorkGroup:DS_Default                       N | 16.7 hr   |   0.4 sec |       0  
State O                                                                  O<100%
Last  O                                                                  O<100%
Wait :                OTHER 100% | Compatibility Mode % ..............<    0%
Queue/Priority     :   D202      | Native Mode Switches ..............<    0/sec
Scheduling State   : Long Wait   | Comp. Mode Switches  ..............<    0/sec
Transactions/min   :   0         | Response Time        ..............<   0.00


Note in particular "Wait: OTHER 100%"

I'm at a loss to determine what "other" thing this is waiting upon

[and as it causes new connections to time out, it is rather frustrating to our users]

* To join/leave the list, search archives, change list settings, *
* etc., please visit  http://www.**--****.com/ *


Re: ODBC listener gone off to never-never land?

Postby Lars Appel » Fri, 17 Sep 2004 16:31:33 GMT




You might try to look at a stack trace in Glance, when
viewing the PIN details. It is function key F3 or similar
and gives you a clue what intrinsic call it might be just
waiting for (probably something socket related).

You could also have a look at the ODBCLOG file (or even
the job listing), but might need to first enable some
logging details. I don't recall the available options
there, but they're most likely documented in the ODBCSE
manual at docs.hp.com ;-)

Lars.

* To join/leave the list, search archives, change list settings, *
* etc., please visit  http://www.**--****.com/ *


Re: ODBC listener gone off to never-never land?

Postby Mike Hornsby » Fri, 17 Sep 2004 22:16:34 GMT

If your trace matches the one below, then this is the normal wait state for
ODBCLNSE. It is sitting on a call to IPCRECVCN waiting for a client to
connect. The IPCRECVCN (from trace) intrinsic allows a server process to
receive a socket connection request and then establish a connection called a
virtual circuit (VC). The connection is identified by the returned VC socket
descriptor. The calling process can then employ the IPCSEND and IPCRECV
intrinsics to send and receive data on the connection.

To see the socket details on this you would use the sockinfo utility.

Cordially,

Mike Hornsby
CoFounder
Beechglen Development Inc. (beechglen.com)
513-922-0509
 XXXX@XXXXX.COM 



From Glance:
NMPRG ODBCLNSE.ODBCSE.SYS
Wait : OTHER

Procedure Trace for Pin 94 is:

PC=a.0018799c enable_int+$2c
NM* 0) SP=41856b30 RP=a.0077756c
notify_dispatcher.block_current_process+$338
NM 1) SP=41856b30 RP=a.007793ac notify_dispatcher+$268
NM 2) SP=41856ab0 RP=a.008682c0 ipc_impede+$268
NM 3) SP=418569b0 RP=a.00868044 ?ipc_impede+$8
export stub: a.02012304 sk_block+$1c8
NM 4) SP=41856870 RP=a.02012c7c sk_block_for_completion+$110
NM 5) SP=41856730 RP=a.02044340 sk_accept.wait_for_passive_open_request+$84
NM 6) SP=418565b0 RP=a.02044644 sk_accept+$200
NM 7) SP=418564f0 RP=a.01ff4be0 IPCRECVCN+$2c0
NM 8) SP=418563f0 RP=a.01ff490c ?IPCRECVCN+$8
export stub: 6d2.00012d1c
NM 9) SP=418561b0 RP=6d2.00024d18
NM a) SP=41855630 RP=6d2.000109b4
NM b) SP=418541b0 RP=6d2.00000000

From Sockinfo:

PROCESS DISPLAY  Pin=94   Lsi=$ea.00000bf0  Dest=0 ($.00000000)       8:55
am
            < #j595 / odbclnse,manager.sys / odbclnse.odbcse.sys >

Desc  Type Prot Addr            Num out     Num in IO Flags
----------------------------------------------------------------------------
---
  15* call tcp  21246                 0          0 i  ns,iw


Socket type: call   Protocol: tcp
Flags: ns,iw
Input count: 1   Output count: 0   Iocount: 1
Socket nwtm pointer: $e6.00001410                    Plfd addr: $a.d5498d94
Read  Enabled mask: $60240000 (close-rply,open/connect,timer,irrec-err)
Write Enabled mask: $00000000
Pending mask: $00000000
Pm pointer: $87772260   Pm id: $00000000
State: listening
Cncts initiated: 0   Cncts received: 0
NetIPC error: 0   BSD error: 0   Pm error: 0
Error location: $0000   Error status: 0
Receive timeout: disabled
Source address: 21246

* To join/leave the list, search archives, change list settings, *
* etc., please visit  http://www.**--****.com/ *


Similar Threads:

1.ODBC listener job stream error

Hi all,

Quick question I hope. I have a the ODBCLNSE job in my startup file
for ages. I've been trying to compile some SQL programs and I'm
getting an access error.
When I check the ODBCLOG I'm finding the following:

1     2006/09/21 12:37:25 0113 Bad run-time args
    2     2006/09/21 12:37:25 0113 Program Abort at
ODBCHOST.C.ODBCLINK(1353)
    3     107/12/28 22:03:16 0076 SE Server must have ODBC_SOCKET_ID
in the rang
e 21246,21248

I tried adding in:

setvar ODBC_SOCKET_ID 21246

to the odbcjob.odbcse before I streamed it to no avail. Can someone
point me to a document that shows how to set this up, or where the
value for the socket Id is set? This was running way before I got
here, so I never went through the initial setup.

Thanks.

Dan

2.HP selling off Compaq land

3.OT: Cleaning your PC Was: [HP3000-L] HP selling off Compaq land

4.OfficeJet G95 doesnt go off-hook

Hi,
My OfficeJet G95 FAX won't go off hook before sending dial-digits, or answer
the phone.

I've checked everything - the line, etc.  The machine rings when called, but
does not pick up.

Short of replacing hardware, does anyone have any suggestions?

I've spent $32 on HP assistance, to no avail.

TIA



5.JetDirect 170x Goes Off Network

Twice now I've had my JetDirect 170x disappear from the network. It
flashes the status light indicating that no network is found.
Disconnect / reconnect of network cable and power don't help.

I printed the JetDirect settings to the printer to find that it was
unable to find its dhcp server. The dhcp server was running fine and
giving out addresses to other devices.

I ran Ethereal and found that there was no traffice whatsoever coming
from the JetDirect. I changed the network cable; no luck.

Finally I power-cycled the hub that it connects to. That fixed it. Now
this is the same hub that supports 2 other pc's, neither of which have
any issue. Both pc's talk 100 MBit; the JetDirect is only 10. I wonder
if there is an issue of it needing to renegotiate its network speed
with the hub? The hub is an SMC 5604DS (which was running fine and had
no error conditions).

Note -- if you are replying to this message, make sure to remove the
NOSPAM from my address.

6. How do you get the Attention light to go off

7. OT - Mt. St. Helens goes off!!

8. HP50G: Emulator hangs when alarm goes off



Return to hp

 

Who is online

Users browsing this forum: No registered users and 3 guest