Baanboard.com

Go Back   Baanboard.com > Forum > Baan Quick Support: Functional & Technical > Operating Systems & Databases

User login

Frontpage Sponsor

Main

Poll
For ERP LN feature pack upgrade, what method of install are you using?
Installation Wizard into existing VRC
37%
Installation Wizard into new VRC
39%
Manual into existing VRC
3%
Manual into new VRC
21%
Total votes: 38

Baanboard at LinkedIn


Reference Content

Reply
 
Thread Tools Display Modes
  #1  
Old 9th July 2002, 21:28
Francesco's Avatar
Francesco Francesco is offline
Guru
 
Join Date: Aug 2001
Location: Antwerp, BE
Posts: 727
Francesco is on a distinguished road
Baan: 5b - DB: Oracle - OS: Solaris
Question Detected database server termination (help! aide! ayuda! hilfe! ayuto! hellup! •‚¯!)!

Baan 5.0b; Solaris2.6; Oracle8.0.5.2

For a long time I have been plagued by mysterious database disconnects (well, that is the error mesg anyway) that seem to have no apparent cause, but happen consistently for periods of time (day, half a day) to certain users running certain sessions.

To make things more complicated, these users and sessions are subject to change.

There are no log messages other than a meaningless one in log.bshell that reads like this:

2002-07-09[09:59:38]:E:knash: ******* S T A R T of Error message *******
2002-07-09[09:59:38]:E:knash: Log message called from /port.6.2a.03.03/vobs/tt/mir/mir/main.c: #1482 keyword: signal handler
2002-07-09[09:59:38]:E:knash: Pid 14321 Uid 3979 Euid 3979 Gid 125 Egid 125
2002-07-09[09:59:38]:E:knash: user_type N language 2 user_name knash tty ote locale ISO88591/NULL
2002-07-09[09:59:38]:E:knash: Errno 0 bdb_errno 0
2002-07-09[09:59:38]:E:knash: Log_mesg: Detected database server termination
2002-07-09[09:59:38]:E:knash: ******* E N D of Error message *******
2002-07-09[09:59:43]:E:knash:
2002-07-09[09:59:43]:E:knash: ******* S T A R T of Error message *******
2002-07-09[09:59:43]:E:knash: Log message called from /port.6.2a.03.03/vobs/tt/lib/nw_1/ipc_fdio.c: #207 keyword: IPC
2002-07-09[09:59:43]:E:knash: Pid 14321 Uid 3979 Euid 3979 Gid 125 Egid 125
2002-07-09[09:59:43]:E:knash: user_type N language 2 user_name knash tty ote locale ISO88591/NULL
2002-07-09[09:59:43]:E:knash: Errno 32 (Broken pipe) bdb_errno 0
2002-07-09[09:59:43]:E:knash: Log_mesg: Connection to server lost: fd_write 14: num_bytes -1 errno 32
2002-07-09[09:59:43]:E:knash: ******* E N D of Error message *******
2002-07-09[09:59:43]:E:knash:
2002-07-09[09:59:43]:E:knash: ******* S T A R T of Error message *******
2002-07-09[09:59:43]:E:knash: Log message called from /port.6.2a.03.03/vobs/tt/mir/mir/main.c: #816 keyword: 4GL/3GL Process information:
2002-07-09[09:59:43]:E:knash: Pid 14321 Uid 3979 Euid 3979 Gid 125 Egid 125
2002-07-09[09:59:43]:E:knash: user_type N language 2 user_name knash tty ote locale ISO88591/NULL
2002-07-09[09:59:43]:E:knash: Errno 32 (Broken pipe) bdb_errno 0
2002-07-09[09:59:43]:E:knash: Active session: whinh3520m000
2002-07-09[09:59:43]:E:knash: Company number: 100
2002-07-09[09:59:43]:E:knash: Executing 3GL function: sql.exec
2002-07-09[09:59:43]:E:knash: Stack information:
2002-07-09[09:59:43]:E:knash: Object Name: otdpurdll4101
2002-07-09[09:59:43]:E:knash: 4GL/3GL Stack Trace
2002-07-09[09:59:43]:E:knash: () (object otdpurdll4101)
2002-07-09[09:59:43]:E:knash: tdpur.dll4101.update.receipts() (object otdpurdll4101)
2002-07-09[09:59:43]:E:knash: tdint.dlltdpur.update.receipts() (object otdintdlltdpur)
2002-07-09[09:59:43]:E:knash: whint.dlltdpur.update.receipts() (object owhintdlltdpur)
2002-07-09[09:59:43]:E:knash: whinh.dll1001.process.receipt() (object owhinhdll1001)
2002-07-09[09:59:43]:E:knash: whinh.dll2210.confirm.receipt.line() (object owhinhdll2210)
2002-07-09[09:59:43]:E:knash: () (object owhinh3520)
2002-07-09[09:59:43]:E:knash: option.confirm.receipt.li() (object owhinh3520)
2002-07-09[09:59:43]:E:knash: () (object ottstpstandard)
2002-07-09[09:59:43]:E:knash: () (object ottstpstandard)
2002-07-09[09:59:43]:E:knash: () (object ottstpstandard)
2002-07-09[09:59:43]:E:knash: std_main() (object ottstpstandard)
2002-07-09[09:59:43]:E:knash: () (object owhinh3520)
2002-07-09[09:59:43]:E:knash:
2002-07-09[09:59:43]:E:knash: ******* E N D of Error message *******
2002-07-09[09:59:43]:E:knash:
2002-07-09[09:59:43]:E:knash: ******* S T A R T of Error message *******
2002-07-09[09:59:43]:E:knash: Log message called from /port.6.2a.03.03/vobs/tt/lib/nw_1/ipc_fdio.c: #207 keyword: IPC
2002-07-09[09:59:43]:E:knash: Pid 14321 Uid 3979 Euid 3979 Gid 125 Egid 125
2002-07-09[09:59:43]:E:knash: user_type N language 2 user_name knash tty ote locale ISO88591/NULL
2002-07-09[09:59:43]:E:knash: Errno 32 (Broken pipe) bdb_errno 0
2002-07-09[09:59:43]:E:knash: Log_mesg: Connection to server lost: fd_write 14: num_bytes -1 errno 32
2002-07-09[09:59:43]:E:knash: ******* E N D of Error message *******

There are no messages on the Oracle side at all, which leads me to believe that Oracle is not a player in this game.
I (and others) have tried numerous things, but nothing even gives me so much as a clue.

My latest suspect is TCP/IP. I have been monitoring netstat, and although I have seen some things that are wrong, it doesn't match (timewise) with the disconnects.

Here is a snapshot taken this morning:

TCP tcpRtoAlgorithm = 4 tcpRtoMin = 1500
tcpRtoMax = 60000 tcpMaxConn = -1
tcpActiveOpens =2106532 tcpPassiveOpens =2084192
tcpAttemptFails = 21490 tcpEstabResets = 56479
tcpCurrEstab = 377 tcpOutSegs =481688864
tcpOutDataSegs =433912860 tcpOutDataBytes =1077445497
tcpRetransSegs =337025 tcpRetransBytes =147782735
tcpOutAck =47745735 tcpOutAckDelayed =26126340
tcpOutUrg = 342 tcpOutWinUpdate = 1807
tcpOutWinProbe = 18416 tcpOutControl =8624900
tcpOutRsts =223745 tcpOutFastRetrans = 38596
tcpInSegs =381734382
tcpInAckSegs =287733539 tcpInAckBytes =1087144468
tcpInDupAck =6731944 tcpInAckUnsent = 0
tcpInInorderSegs =173897177 tcpInInorderBytes =1531281550
tcpInUnorderSegs = 30904 tcpInUnorderBytes =1211479
tcpInDupSegs =487665 tcpInDupBytes =15999061
tcpInPartDupSegs = 9 tcpInPartDupBytes = 6179
tcpInPastWinSegs = 39 tcpInPastWinBytes =3718206987
tcpInWinProbe = 35 tcpInWinUpdate = 17320
tcpInClosed = 8221 tcpRttNoUpdate =113036
tcpRttUpdate =285120798 tcpTimRetrans =264558
tcpTimRetransDrop = 16121 tcpTimKeepalive = 30564
tcpTimKeepaliveProbe= 12212 tcpTimKeepaliveDrop = 36
tcpListenDrop = 379 tcpListenDropQ0 = 0
tcpHalfOpenDrop = 0

IP ipForwarding = 2 ipDefaultTTL = 255
ipInReceives =320519386 ipInHdrErrors = 0
ipInAddrErrors = 0 ipInCksumErrs = 0
ipForwDatagrams = 0 ipForwProhibits = 0
ipInUnknownProtos =1183533 ipInDiscards = 0
ipInDelivers =748178932 ipOutRequests =417736747
ipOutDiscards = 0 ipOutNoRoutes = 0
ipReasmTimeout = 60 ipReasmReqds = 0
ipReasmOKs = 0 ipReasmFails = 0
ipReasmDuplicates = 0 ipReasmPartDups = 0
ipFragOKs = 0 ipFragFails = 0
ipFragCreates = 0 ipRoutingDiscards = 0
tcpInErrs = 2646 udpNoPorts =11692887
udpInCksumErrs = 1 udpInOverflows = 0
rawipInOverflows = 0


Finally, it seems that bouncing Baan has a positive effect on the problem. This also puzzles me, because I can't make the connection there either.

In short, I am totally lost and if I bang my head against the wall one more time, it might give in (either the wall or my head).

Any suggestions?
__________________
Cheers,

Francesco
..............................................................

Admiral Business Solutions | My World | Baan Board | IT Happens!

"If everyone is thinking alike, then somebody isn't thinking" -- George Patton
"It's easy to cry 'bug' when the truth is that you've got a complex system and sometimes it takes a while to get all the components to co-exist peacefully." -- Doug Vargas
Reply With Quote
  #2  
Old 10th July 2002, 11:24
Djie-En's Avatar
Djie-En Djie-En is offline
Senior Member
 
Join Date: Apr 2002
Location: The Netherlands
Posts: 197
Djie-En is on a distinguished road
Baan: BaanIVc4 - DB: Oracle11.2 - OS: AIX6.1
Disconnection

Hi,

Is it possible that the ora_timeout (db_resource file) has disconnect the user?
Is there nothing in the ALERT-file?
Or in the ERRPT on the server?
Are there other programs running, such as Closeidle or something like that?

GN
Reply With Quote
  #3  
Old 10th July 2002, 19:07
Francesco's Avatar
Francesco Francesco is offline
Guru
 
Join Date: Aug 2001
Location: Antwerp, BE
Posts: 727
Francesco is on a distinguished road
Baan: 5b - DB: Oracle - OS: Solaris
Djie,

It is certainly possible that ora_timeout is disconnecting the user.
I have a rather generous setting in db_resource (at least I think so, but if anybody disagrees I'd like to hear it):
ora_timeout:{6000,300,300,300,300}
If it is indeed timing out, the question is ... why is it timing out? I also thought that this would create an entry in log.oracle but I am not sure there.

Oracle's alert files are coming up daisies just like the Baan logs.

ERRPT is (if I remember correctly) an AIX command, that to my knowledge has no equivalent on Solaris.

There are no third party programs involved.
__________________
Cheers,

Francesco
..............................................................

Admiral Business Solutions | My World | Baan Board | IT Happens!

"If everyone is thinking alike, then somebody isn't thinking" -- George Patton
"It's easy to cry 'bug' when the truth is that you've got a complex system and sometimes it takes a while to get all the components to co-exist peacefully." -- Doug Vargas
Reply With Quote
  #4  
Old 10th July 2002, 21:55
victor_cleto's Avatar
victor_cleto victor_cleto is offline
Guru
 
Join Date: Aug 2001
Location: Portugal
Posts: 776
victor_cleto is on a distinguished road
Baan: none (B40c4 was last) - DB: Oracle - OS: Linux (RHEL)
Some ideas

Use a user as test with debug information related to the Baan DB driver. You don't need to have it doing a lot (so that the log does not increase a lot, or create a script that keeps moving the log and compressing it to avoid running out of space), just enough work, time logged in to be able to get some information from the Db driver related to the drop of the connection.

This may give you a clue for the disconnection.
The timeout seems OK (we use 0=none), and if a user is not doing anything then that should not be a problem.

Also, check your lan settings, we experienced performance problems once (and this may cause disconnections) when one card was in full duplex and another was in half-duplex, worth to have your network checked and look for a high count of collisions.

What about kernel settings?!
Reply With Quote
Sponsored Links
  #5  
Old 10th July 2002, 22:27
Francesco's Avatar
Francesco Francesco is offline
Guru
 
Join Date: Aug 2001
Location: Antwerp, BE
Posts: 727
Francesco is on a distinguished road
Baan: 5b - DB: Oracle - OS: Solaris
Those are all good suggestions Victor.

I actually created a DBLOG file, when I was able to reproduce the error some time ago. The log was terminated as abruptly as the connection. I was unable to extract any valuable info from it.
Here are the last 10 lines...it simply stops after this.

Bind :1 : string : '040'
Bind :2 : string : ' 51-'
------ QPS Output Row -------
Bind 1 : blck : char : <2>
Bind 2 : dsca : string : 'FRESNO PID MFG '
Bind 3 : ffno : long : <452>
Bind 4 : ffst : char : <0>
Bind 5 : nrgr : string : '040'
Bind 6 : seri : string : ' 51-'
Bind 7 : lngt : char : <3>


LAN (WAN) and performance are high on my suspect list. I did ask this one user to try it from a different machine however, and he got the same result. What's worse, he logged on as a different user and got the same result. That user himself could work just fine from ahy machine.
Now if it was just that one user, I would probably see what's in his desk drawers before I see what's in his PC, but it happens to a handful of users across the country.
Same users, same error, comes and goes. :confused:

I did check for network collissions, but everything looks clean.
example:
Name Mtu Net/Dest Address Ipkts Ierrs Opkts Oerrs Collis Queue
lo0 8232 loopback localhost 446064368 0 446064368 0 0 0
qfe0 1500 otherserver otherserver 44215519 0 33175362 0 0 0
qfe5 1500 myserver myserver 362317672 7 396039668 0 0 0


I had a team of hotshot IBM consultants check out my kernel settings last year, and they were able to solve most (if not all) of the performance problems we were suffering then.
It is probably time to do evaluate these settings again (more sites, more users, more everything), but somehow I don't think that this would be the core problem.
Then again, if I knew what was at the core of this problem I would have bloody well fixed it by now, right? ;)
So here you go (sar -v 5 5):

SunOS uscobrmfa-ue-25 5.6 Generic_105181-26 sun4u 07/10/02

13:21:55 proc-sz ov inod-sz ov file-sz ov lock-sz
13:22:00 1682/16394 0 31168/69992 0 22043/22043 0 0/0
13:22:05 1684/16394 0 31168/69992 0 22056/22056 0 0/0
13:22:10 1681/16394 0 31168/69992 0 22034/22034 0 0/0
13:22:15 1678/16394 0 31168/69992 0 22016/22016 0 0/0
13:22:20 1670/16394 0 31168/69992 0 21954/21954 0 0/0


etc/system :

set scsi_options=0x7f8
set sd:sd_io_time=0x78
set sd:sd_max_throttle=20
set maxphys=2097152

set tune_t_gpgslo=250
set tune_t_minarmem=100
set tune_t_minasmem=250
set rstchown=0

* Message Queue Parameters
set msgsys:msginfo_msgmap=712
set msgsys:msginfo_msgmax=8192
set msgsys:msginfo_msgmnb=65536
set msgsys:msginfo_msgmni=1024
set msgsys:msginfo_msgseg=8192
set msgsys:msginfo_msgssz=16
set msgsys:msginfo_msgtql=1802

* Semaphore Parameters
set semsys:seminfo_semmap=3202
set semsys:seminfo_semmni=3200
set semsys:seminfo_semmns=3200
set semsys:seminfo_semmnu=3200
set semsys:seminfo_semmsl=3200
set semsys:seminfo_semopm=32
set semsys:seminfo_semume=100

* Shared Memory Parameters
set shmsys:shminfo_shmmax=0xffffffff
set shmsys:shminfo_shmmni=320
set shmsys:shminfo_shmseg=200

set priority_paging=1

set pln:pln_enable_detach_suspend=1
set soc:soc_enable_detach_suspend=1
set socal:socal_enable_suspend=1
set tcp:tcp_conn_hash_size=8192
set dosynctodr=0

* vxfs_START -- do not remove the following lines:
set lwp_default_stksize=0x4000
set rpcmod:svc_run_stksize=0x4000
* vxfs_END

* vxvm_START (do not remove)
forceload: drv/atf
forceload: drv/pln
forceload: drv/ses
forceload: drv/vxdmp
forceload: drv/vxio
forceload: drv/vxspec
forceload: drv/sd
forceload: drv/QLGC,isp
forceload: drv/sbus
forceload: drv/ssd
forceload: drv/sf
forceload: drv/soc
forceload: drv/socal
rootdev:/pseudo/vxio@0:0
set vxio:vol_rootdev_is_volume=1
* vxvm_END (do not remove)

forceload: drv/st
forceload: drv/sg
forceload: drv/fcaw

set rlim_fd_cur=2048
set rlim_fd_max=8192
set npty=350

set ecache_calls_a_sec=1
set ecache_scan_rate=1
*set ecache_is_mirrored=1

* force all qfe interfaces to 100MB Full Duplex No Autoneg
set qfe:qfe_adv_autoneg_cap=0
set qfe:qfe_adv_100T4_cap=0
set qfe:qfe_adv_100fdx_cap=1
set qfe:qfe_adv_100hdx_cap=0
set qfe:qfe_adv_10fdx_cap=0
set qfe:qfe_adv_10hdx_cap=0


See anything disturbing?
__________________
Cheers,

Francesco
..............................................................

Admiral Business Solutions | My World | Baan Board | IT Happens!

"If everyone is thinking alike, then somebody isn't thinking" -- George Patton
"It's easy to cry 'bug' when the truth is that you've got a complex system and sometimes it takes a while to get all the components to co-exist peacefully." -- Doug Vargas
Reply With Quote
  #6  
Old 10th July 2002, 23:14
patvdv's Avatar
patvdv patvdv is offline
Board Master
 
Join Date: Aug 2001
Location: Belgium
Posts: 2,167
patvdv is on a distinguished road
Baan: n/a - DB: n/a - OS: AIX, HP-UX, Linux
Patterns

Hi Francesco,

These are tough problems to face. Are there any patterns you can see in the a series of users getting disconnected? Ie: locations, time of day, type of activities running, software configurations etc?

I am not very familiar with Solaris kernels but it looks OK to me. The network bit can be harder to diagnose. You wouldn't only have to check locally on the your system but also on your routers/switches, the performance of your WAN etc. I would definatively get some networking people involved in all of this and possibly ask your WAN provider to pay some extra attention to your WAN's health. Have you seen any strange events happening on your local network?

Has the Oracle listener log files revealed anything? You could consider switching the listener into debug or support level mode and see if that would yield anything. Problems like these are usually a matter of eliminating possibilities...
__________________
Regards,

Patrick Van der Veken - Admin & Founder - (c) 2001-2017 baanboard.com/baanforums.com
Reply With Quote
  #7  
Old 10th July 2002, 23:43
Francesco's Avatar
Francesco Francesco is offline
Guru
 
Join Date: Aug 2001
Location: Antwerp, BE
Posts: 727
Francesco is on a distinguished road
Baan: 5b - DB: Oracle - OS: Solaris
Tough problems indeed, Pat

There is no obvious similarity between the users. There are basically three sessions, whinh3520m000, tipcf5530m000 and tisfc0101s000.
I had our DBA look into the related tables and he assures me that there is nothing wrong with them on the Oracle side.
Timewise there is once again no indication. It comes and goes in a seamingly random pattern. If there are any related processes, they don't jump out.

I think I can eliminate switches, because if they would be the problem I would see the disconnects on a larger scale for certain sites.

I am definitely going to get some network people involved, but as always, in order to get the right answer, I will have to ask the right question and I am not so sure what that is at this point.

There have been some strange events yes, but I don't know how they would relate.
Our NIS password file got corrupted causing some users to log on as root (or attempt to).
Then there is the snapshot I took from the TCP data, where the value of tcpListenDrop is bothering me. This indicates that we are operating with a full TCP stack, which apparently occurs from time to time. However, not all the time and once again it doesn't seem to coinside with the database disconnects.
The tcpInunOrderBytes also seems high to me. If I remember correctly this indicates a routing problem.
Unfortunately this is too far out of my area to make an educated diagnosis.

As you can see, I am walking the long road of elimination but slowly running out of options.
I appreciate all you guys' help so far. Even if you don't provide a solution, it helps me to retrace my steps and rethink my thoughts. The truth is out there (somewhere).

I'll run your Oracle suggestions by the DBA. Problem is once again that I don't think anything is actually happening on the Oracle side or there would be some entry (any entry) in the logs.

I think I'll sign up for one of Jim's performance classes and 'allow' him to use this as a casus.
__________________
Cheers,

Francesco
..............................................................

Admiral Business Solutions | My World | Baan Board | IT Happens!

"If everyone is thinking alike, then somebody isn't thinking" -- George Patton
"It's easy to cry 'bug' when the truth is that you've got a complex system and sometimes it takes a while to get all the components to co-exist peacefully." -- Doug Vargas
Reply With Quote
  #8  
Old 11th July 2002, 00:30
Francesco's Avatar
Francesco Francesco is offline
Guru
 
Join Date: Aug 2001
Location: Antwerp, BE
Posts: 727
Francesco is on a distinguished road
Baan: 5b - DB: Oracle - OS: Solaris
DUH!

Did I mention we are running in host mode?

I was typing out an email to the DBA about the listener log when it finally hit me.....no such thing.
__________________
Cheers,

Francesco
..............................................................

Admiral Business Solutions | My World | Baan Board | IT Happens!

"If everyone is thinking alike, then somebody isn't thinking" -- George Patton
"It's easy to cry 'bug' when the truth is that you've got a complex system and sometimes it takes a while to get all the components to co-exist peacefully." -- Doug Vargas
Reply With Quote
  #9  
Old 11th July 2002, 00:33
patvdv's Avatar
patvdv patvdv is offline
Board Master
 
Join Date: Aug 2001
Location: Belgium
Posts: 2,167
patvdv is on a distinguished road
Baan: n/a - DB: n/a - OS: AIX, HP-UX, Linux
Customized?

LOL Francesco, no you didn't so skip the listener thing. Maybe something else to keep in mind: any session customizations?
__________________
Regards,

Patrick Van der Veken - Admin & Founder - (c) 2001-2017 baanboard.com/baanforums.com
Reply With Quote
  #10  
Old 11th July 2002, 18:16
ssellens's Avatar
ssellens ssellens is offline
Member
 
Join Date: Oct 2001
Posts: 44
ssellens is on a distinguished road
Baan: BaanIVc4 - DB: Oracle8i Oracle 9i - OS: Solaris 8,9 & 10
Hi Francesco,

Are they all happening in the morning? maybe when the user is first using the Baan GUI? I ask because we just had a user complain about the same kind of problem, however it turns out that she is trying to use the GUI that she has left it running overnight. This means after the overnight back-up when Oracle is shutdown/restarted.

USERS!!!!

Not sure it is the same, but thought I would mention it.

Best of luck

Steve
__________________
Stephen J. Sellens
Unix/Baan/Oracle Administrator
Reply With Quote
  #11  
Old 11th July 2002, 18:35
Francesco's Avatar
Francesco Francesco is offline
Guru
 
Join Date: Aug 2001
Location: Antwerp, BE
Posts: 727
Francesco is on a distinguished road
Baan: 5b - DB: Oracle - OS: Solaris
Thanks Steve

I often say that my job would be a lot easier without 'them users'.

I've had it happen mornings, afternoons and alldaylongs (even at night, come to think of it, when I am the only idiot on the system. What would indicate that (network)performance maybe is not an issue after all. After I finally gave up in frustration around 3am, I found an email the next morning thanking me for fixing the problem. Of course I took full credit )

We run hot back-ups of Oracle only, and I have a small perl script that sends users who are logged on at 2am a logoff reminder and me a list of who was online. No match.
__________________
Cheers,

Francesco
..............................................................

Admiral Business Solutions | My World | Baan Board | IT Happens!

"If everyone is thinking alike, then somebody isn't thinking" -- George Patton
"It's easy to cry 'bug' when the truth is that you've got a complex system and sometimes it takes a while to get all the components to co-exist peacefully." -- Doug Vargas
Reply With Quote
  #12  
Old 11th July 2002, 20:31
victor_cleto's Avatar
victor_cleto victor_cleto is offline
Guru
 
Join Date: Aug 2001
Location: Portugal
Posts: 776
victor_cleto is on a distinguished road
Baan: none (B40c4 was last) - DB: Oracle - OS: Linux (RHEL)
Strange that you mention that it mostly happens with some sessions that others... You should have them checked, and the sql statements they issue then checked by your DBA: not always what you see from the DB point of view correspond to what is optimal from Baan point of view, and vice-versa!

Also, may worth to have checked the portingset (more like the db driver itself, but you never know, you may encounter there the explanation/solution to your problem).
Reply With Quote
  #13  
Old 11th July 2002, 20:51
Francesco's Avatar
Francesco Francesco is offline
Guru
 
Join Date: Aug 2001
Location: Antwerp, BE
Posts: 727
Francesco is on a distinguished road
Baan: 5b - DB: Oracle - OS: Solaris
Unfortunately we are stuck in our current porting set, which is a different story and I believe I ranted about that already in another thread.

I agree, the fact that it is limited to some sessions is suspicious.

This morning I was being hammered again with this error. I am in the process of creating a variety of dbslogs (again, just for the heck of it), and while doing so, I noticed that the session I am logging (tisfc0101s000, production orders), is reading half the DD as it is loading.

This is something I didn't consider, I thought entering a production order would be a relative simple, or isolated, task.

Now NPRao mentioned in pm that he had seen something similar with a bad file in the DD.
I discarded this possibility because it didn't match the fact that it happens only to certain users.

however, now seeing how much of the DD this session uses, I am focussing my attention in that direction again.
What's more, as I mentioned, bouncing the app seems to have a positive effect.

*drumroll*

Could it be that this is a result of corrupted shared memory? But then again, that does not explain why the problem comes and goes without restarting Baan.

My head...your wall...let's get it on.
__________________
Cheers,

Francesco
..............................................................

Admiral Business Solutions | My World | Baan Board | IT Happens!

"If everyone is thinking alike, then somebody isn't thinking" -- George Patton
"It's easy to cry 'bug' when the truth is that you've got a complex system and sometimes it takes a while to get all the components to co-exist peacefully." -- Doug Vargas
Reply With Quote
  #14  
Old 12th July 2002, 00:01
Francesco's Avatar
Francesco Francesco is offline
Guru
 
Join Date: Aug 2001
Location: Antwerp, BE
Posts: 727
Francesco is on a distinguished road
Baan: 5b - DB: Oracle - OS: Solaris
yep, shared memory

Here is what I got as I was creating a dbslog with 0001111 options on inserting a record using tisfc0101s000 (production orders):

(fragment)


<5327> ----- Dictionary for tccom130 compnr 100 -----

Physical table name : 'baan.ttccom130100'
Owner class : group
Number of total columns : 27
Number of real columns : 27
Number of hash columns : 0
Number of pseudo columns : 0
Number of base columns : 27
Number of combined columns : 5
Number of column relations : 3
Number of column conversions : 0
Number of indexes : 4
Refresh Time: 5
Table idxopt: 0214
n Column Type Size Dept Elem Flag Digv Diga DbName
1 cadr 6 9 0 0 12608 0 0 t$cadr
2 nama 6 35 0 0 4160 0 0 t$nama
3 namb 6 30 0 0 4160 0 0 t$namb
4 namc 6 30 0 0 4160 0 0 t$namc
5 namd 6 30 0 0 4160 0 0 t$namd
6 hono 6 10 0 0 4160 0 0 t$hono
7 pobn 6 10 0 0 4160 0 0 t$pobn
8 name 6 30 0 0 4160 0 0 t$name
9 namf 6 30 0 0 4160 0 0 t$namf
10 pstc 6 10 0 0 4416 0 0 t$pstc
11 crte 6 5 0 0 4160 0 0 t$crte
12 telp 6 15 0 0 4416 0 0 t$telp
13 telx 6 15 0 0 4160 0 0 t$telx
14 tefx 6 15 0 0 4160 0 0 t$tefx
15 ccty 6 3 0 0 4416 0 0 t$ccty
16 cste 6 2 0 0 4160 0 0 t$cste
17 seak 6 16 0 0 4416 0 0 t$seak
18 fovn 6 20 0 0 4160 0 0 t$fovn
19 lvdt 14 4 0 0 4160 0 0 t$lvdt
20 coaf 6 3 0 0 4160 0 0 t$coaf
21 ccal 6 9 0 0 4160 0 0 t$ccal
22 geoc 6 10 0 0 4160 0 0 t$geoc
23 dtlm 14 4 0 0 4160 0 0 t$dtlm
24 txta 3 4 0 0 4160 10 0 t$txta
25 tzid 6 3 0 0 4160 0 0 t$tzid
26 Refcntd 3 4 0 0 64 0 0 t$Refcntd
27 Refcntu 3 4 0 0 64 0 0 t$Refcntu

Combined columns:
Name Parts
_index1 cadr
_index2 seak cadr
_index3 ccty pstc cadr
_index4 telp cadr
cmba ccty cste

Indexes:
I Dupl Opt Status Parts
1 0 214 1 cadr
2 0 214 1 seak cadr
3 0 214 1 ccty pstc cadr
4 0 214 1 telp cadr
--------------------------------

dbs_handle_ddinfo done.
Store dbs_dd in shared memory, key = 'tccom130100'
Msg_type 11 received.
Search DBS_DD in shared memory for 'tdisa001100' - Found but invalidated
Searching for DD with other compnr: not found
Search DBS_DD in shared memory for 'tttxt001100' - Found but invalidated
Searching for DD with other compnr: not found
Search DBS_DD in shared memory for 'tdisa001100' - Found but invalidated
Searching for DD with other compnr: not found
Search DBS_DD in shared memory for 'tttxt001100' - Found but invalidated
Searching for DD with other compnr: not found
Search DBS_DD in shared memory for 'tttxt001100' - Found but invalidated
Searching for DD with other compnr: not found
Search DBS_DD in shared memory for 'tttxt001100' - Found but invalidated
Searching for DD with other compnr: not found
Search DBS_DD in shared memory for 'tttxt001100' - Found but invalidated
Searching for DD with other compnr: not found
Search DBS_DD in shared memory for 'tttxt001100' - Found but invalidated
Searching for DD with other compnr: not found
Search DBS_DD in shared memory for 'tdipu001100' - Found but invalidated
Searching for DD with other compnr: not found
Search DBS_DD in shared memory for 'tttxt001100' - Found but invalidated
Searching for DD with other compnr: not found
Search DBS_DD in shared memory for 'tttxt001100' - Found but invalidated
Searching for DD with other compnr: not found
Search DBS_DD in shared memory for 'tttxt001100' - Found but invalidated
Searching for DD with other compnr: not found
Search DBS_DD in shared memory for 'tdipu001100' - Found but invalidated
Searching for DD with other compnr: not found
Search DBS_DD in shared memory for 'tttxt001100' - Found but invalidated
Searching for DD with other compnr: not found


THen the session continues to spin on those two tables. I had to kill it after several hours.

Looks like we have a winner.
__________________
Cheers,

Francesco
..............................................................

Admiral Business Solutions | My World | Baan Board | IT Happens!

"If everyone is thinking alike, then somebody isn't thinking" -- George Patton
"It's easy to cry 'bug' when the truth is that you've got a complex system and sometimes it takes a while to get all the components to co-exist peacefully." -- Doug Vargas
Reply With Quote
  #15  
Old 19th July 2002, 17:13
Francesco's Avatar
Francesco Francesco is offline
Guru
 
Join Date: Aug 2001
Location: Antwerp, BE
Posts: 727
Francesco is on a distinguished road
Baan: 5b - DB: Oracle - OS: Solaris
Update

We've been 'clean' for a week now as far as the 'detected database server disconnects' go.

My initial attempt to bounce Baan did not work though. Wether all SM attachments weren't released or the corruption laid in the DD itself I never found out. I rebuilt the DD just in case and made sure Baan bounced clean by stopping and starting the app manually.

This seems to have done the trick.

Thanks for your help everyone. Kudos to NPRao. He actually pm'd me this solution right after my initial post, but I guess I _had_ to take the long way home.
__________________
Cheers,

Francesco
..............................................................

Admiral Business Solutions | My World | Baan Board | IT Happens!

"If everyone is thinking alike, then somebody isn't thinking" -- George Patton
"It's easy to cry 'bug' when the truth is that you've got a complex system and sometimes it takes a while to get all the components to co-exist peacefully." -- Doug Vargas
Reply With Quote
Reply


Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is Off
HTML code is Off
Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
How to split Baan Application and Baan Database ? baaniac Operating Systems & Databases 4 2nd June 2009 04:48
Detected Database Server Termination mprakash Tools Administration & Installation 8 9th August 2004 14:07
Detected database server termination & Data dictionary not found kevinlee Tools Administration & Installation 1 2nd April 2004 23:09


All times are GMT +2. The time now is 15:36.


©2001-2017 - Baanboard.com - Baanforums.com