We have installed porting 6.1c.07.09 on a BaaN IV c4 installation.
Tests are OK but this. Users are logout with a 'Conection with server lost' message and here are lines from log.bshell6.1


2007-08-24[12:01:29]:E:taa: ******* S T A R T of Error message *******
2007-08-24[12:01:29]:E:taa: Log message called from /BAAN/view/port.6.1c.07.09/vobs/tt/mir/mir/main.c: #1676 keyword: signal handler
2007-08-24[12:01:29]:E:taa: Pid 26355 Uid 1012 Euid 1012 Gid 125 Egid 125
2007-08-24[12:01:29]:E:taa: user_type N language 5 user_name taa tty ote locale ISO88591/NULL
2007-08-24[12:01:29]:E:taa: Errno 0 bdb_errno 0
2007-08-24[12:01:29]:E:taa: Log_mesg: Detected database server termination
2007-08-24[12:01:29]:E:taa: ******* E N D of Error message *******
2007-08-24[12:01:34]:E:taa:
2007-08-24[12:01:34]:E:taa: ******* S T A R T of Error message *******
2007-08-24[12:01:34]:E:taa: Log message called from /BAAN/view/port.6.1c.07.09/vobs/tt/lib/nw_1/ipc_fdio.c: #225 keyword: IPC
2007-08-24[12:01:34]:E:taa: Pid 26355 Uid 1012 Euid 1012 Gid 125 Egid 125
2007-08-24[12:01:34]:E:taa: user_type N language 5 user_name taa tty ote locale ISO88591/NULL
2007-08-24[12:01:34]:E:taa: Errno 32 (Broken pipe) bdb_errno 0
2007-08-24[12:01:34]:E:taa: Log_mesg: Connection to server lost: fd_write 12: num_bytes -1 errno 32
2007-08-24[12:01:34]:E:taa: ******* E N D of Error message *******
2007-08-24[12:01:35]:E:taa:
2007-08-24[12:01:35]:E:taa: ******* S T A R T of Error message *******
2007-08-24[12:01:35]:E:taa: Log message called from /BAAN/view/port.6.1c.07.09/vobs/tt/lib/al_1/al_log.c: #1191 keyword: stack trace
2007-08-24[12:01:35]:E:taa: Pid 26355 Uid 1012 Euid 1012 Gid 125 Egid 125
2007-08-24[12:01:35]:E:taa: user_type N language 5 user_name taa tty ote locale ISO88591/NULL
2007-08-24[12:01:35]:E:taa: Errno 2 (No such file or directory) bdb_errno 0
2007-08-24[12:01:35]:E:taa: 26355: bshell (taa@0000-PC-003018:3472/SOCKET) -1 545 0 1
2007-08-24[12:01:35]:E:taa: ff1a03e4 waitid (0, 6f39, ffbfb488, 103)
2007-08-24[12:01:35]:E:taa: ff195124 system (ffbfb758, 1f7488, 66f3, 24b61c, 0, 235000) + 240
2007-08-24[12:01:35]:E:taa: 000e99cc create_stack_trace (1f7400, 24b400, ffbeb924, 235000, 1f7400, 1) + 98
2007-08-24[12:01:35]:E:taa: 000e985c log_stack_trace (0, 6, ffbeb924, 2506a8, 27, 44) + 4
2007-08-24[12:01:35]:E:taa: 000e99e4 crash_notification (0, 4, 1, 0, 183, 21c400) + 4
2007-08-24[12:01:35]:E:taa: 000fe328 do_emergency (24df10, 252c9c, 252cc0, 0, e99e0, 252c00) + 28
2007-08-24[12:01:35]:E:taa: 001a25fc detach_on_io_error (1, e1, 200de0, 0, 24b800, 141) + 48
2007-08-24[12:01:35]:E:taa: 001a1d30 fd_write (200c00, ffffffff, ffffffff, 27b880, 261000, 20) + c4
2007-08-24[12:01:35]:E:taa: 000f8af4 sflush (24df48, 24df48, 2284e4, ffbf9f48, 2548b0, 20) + 54
2007-08-24[12:01:35]:E:taa: 0012ffc4 client_commu (27b838, 131630, 221834, 254898, 14, 5) + 3c
2007-08-24[12:01:35]:E:taa: 0012f730 bdb_do_exec_db (27b8d0, 8005, 5, 2548a0, 254800, 260fa0) + d4
2007-08-24[12:01:35]:E:taa: 001391e8 bdb_exec_db (5, 1, 0, 268c00, 27b4f4, 3) + 5bc
2007-08-24[12:01:35]:E:taa: 00070840 free_all_pr_tables (29b008, 21e440, 260c00, 21eca0, 2d0318, 21ec00) + 74
2007-08-24[12:01:35]:E:taa: 00065038 process_destroy_fase_1 (29b008, 0, 21e400, 138, 279c48, 3) + 90
2007-08-24[12:01:35]:E:taa: 00064f74 process_unlink (29b008, 0, 235400, 21e5d8, 276460, 261800) + 58
2007-08-24[12:01:35]:E:taa: 00121ac8 _DsForHashTable (21db90, 64f1c, 0, 21dbc0, 0, 8) + 34
2007-08-24[12:01:35]:E:taa: 0005e194 stop (ffffffff, 4, 2284e4, ffbf9f48, ffbfaf48, 0) + e4
2007-08-24[12:01:35]:E:taa: 0005ed40 server_death (1f0800, 1, 246c00, 1f0800, 1f0800, 0) + 68
2007-08-24[12:01:35]:E:taa: 001a25cc detach_on_io_error (27b880, 2516a8, 1000, 0, 0, 5ecd8) + 18
2007-08-24[12:01:35]:E:taa: 001a1c50 fd_read (0, 2516a8, 1000, 27b880, 1000, 261000) + c4
2007-08-24[12:01:35]:E:taa: 000f9e14 _sfilbuf (24df10, 2516a8, 9, a03, ce, d91) + fc
2007-08-24[12:01:35]:E:taa: 000fb9a0 r_long (24df10, ffffffff, 26f, ffbfc13c, 0, ffbfc13c) + 24
2007-08-24[12:01:35]:E:taa: 00131280 client_get_commu_sql (254898, 24df10, 2516a8, 260c00, ffff8000, 13127c) + 4
2007-08-24[12:01:35]:E:taa: 0012ffe8 client_commu (27b838, 132488, 221834, 254898, 28, a) + 60
2007-08-24[12:01:35]:E:taa: 0012fc70 bdb_exec_sql (27b838, 8965c0, 816420, 0, 816478, ffbfc2dc) + 60
2007-08-24[12:01:35]:E:taa: 0015dd14 get_db_view_result (816420, 261800, 28, 6534e0, 816408, 2) + 80
2007-08-24[12:01:35]:E:taa: 0015cc08 intern_fetch (8163e0, 1, 815408, 0, 9, 816af0) + 70
2007-08-24[12:01:35]:E:taa: 0015cd74 get_next_for_fetch (815408, 0, 0, 261800, 8163e0, 815510) + a0
2007-08-24[12:01:35]:E:taa: 00146e54 sql_fetch (815408, 3, 811570, 0, 0, 1) + 118
2007-08-24[12:01:35]:E:taa: 00074d4c mir_sql_fetch (40000, 260c00, 815408, 262400, 763558, 260c00) + 3c
2007-08-24[12:01:35]:E:taa: 00041dd0 baan_cpu (260c00, 268c00, 8000, 260c00, ffff8000, 260c00) + 4eb0
2007-08-24[12:01:35]:E:taa: 0005db44 main (83c4, 228400, 216800, 261800, 276460, 21b800) + b48
2007-08-24[12:01:35]:E:taa: 00035888 _start (0, 0, 0, 0, 0, 0) + 108
2007-08-24[12:01:35]:E:taa:
2007-08-24[12:01:35]:E:taa: ******* E N D of Error message *******
2007-08-24[12:01:35]:E:taa:
2007-08-24[12:01:35]:E:taa: ******* S T A R T of Error message *******
2007-08-24[12:01:35]:E:taa: Log message called from /BAAN/view/port.6.1c.07.09/vobs/tt/lib/nw_1/ipc_fdio.c: #225 keyword: IPC
2007-08-24[12:01:35]:E:taa: Pid 26355 Uid 1012 Euid 1012 Gid 125 Egid 125
2007-08-24[12:01:35]:E:taa: user_type N language 5 user_name taa tty ote locale ISO88591/NULL
2007-08-24[12:01:35]:E:taa: Errno 32 (Broken pipe) bdb_errno 0
2007-08-24[12:01:35]:E:taa: Log_mesg: Connection to server lost: fd_write 12: num_bytes -1 errno 32
2007-08-24[12:01:35]:E:taa: ******* E N D of Error message *******


I have searched BaaNboard and most similar is this post (http://www.baanboard.com/baanboard/showthread.php?t=5760&highlight=Connection+server+lost%3A+fd_write)

¿Can anybody help me?
Is there any kind of DD rebuild recommended?, in which order?, for the whole application?

Thank's in advance.

Is it happening at the server side? If not and only happening at the user side then, what BW version are you running at the client side? Is it B40c.80..??

sk

This kind of messages/errors appear after compatibility issues.

Please check your Porting Set release notes. Whether they have suggested to load any tools solution OR compatibility among "OS Version", "Porting Set Version", "BW Client Version" (as suggeted by Sukesh), "BaaN Version", etc.....

All these should be compatible with each other.

Regards,
Kaustubh

was your problem resolved?

Thank you very much for your answers and sorry because I have not replied before.

We still have the problem.

We have tested all compatibility and Porting set notes.
We have opened two cases, in Oracle and Infor.

It appears to be an OCI error. Oracle suggested us to go to last 9.2 patch, so this weekend we migrated database to 9.2.0.8.

Also BW version is 080, we changed ipc_info in order to work with sockets instead pipes, installed also multi process OS patch (Solaris 9).

Everything in one step to make good analysis.

Infor suggested us to make some changes in db_resource introducing max_free_cursor:50 and retained_cursors:40


Well, we are still in the problem. :(

I'll tell you any more, hope good things in next future.
All your help is pretty apreciated.

Is this a new install?

was it working before? if so what changed?

what is the output of ulimit -a on that machine?

Dave

It is not a new install.
We upgraded Porting set, BW client, OS Sun and Oracle.

Actual version for each one is:

PS - 61c.07.09
BW - 080
OS - Sun Solaris 2.9
Oracle 9.2.0.8

Also we have tested sessions in ASCII mode and the same problem.

Sessions that most often produce this problem are:

- Batches finalization process
- Post Integration transactions to Finance transactions
- Process payments

Result for ulimit -a is

time(seconds) unlimited
file(blocks) unlimited
data(kbytes) unlimited
stack(kbytes) 8192
coredump(blocks) unlimited
nofiles(descriptors) 1024
vmemory(kbytes) unlimited

Thank you.

stack is too low try upping it to maybe 40960

Dave

After all, solution was install last porting set version.

Now with 6.1c.07.15 everything is normal.
We have tested it before for two weeks in a test server and now in 'real life', and thinks are fine after two complete days.

Thank you to all for your suggestions.
dave_23, solution you mentioned about changing stack size apply to HP-UX, I think.

SUN told us not to change those kernel parameters.
Infor recommended us follow Sun recomendation.

A solution could be change it for the bshell and child process, but not in a global way. Support recommended us install another Porting and was OK.

So we could not test what you told us.
Thank you too, of course

And we do not know what was exactly the problem. (If was different than a Porting Set version) ;)

check the entry in routing table.

# netstat -r

check the gateway and ip settings , are they valid ?

and also check the host file