Hi!

Sorry for 3.x type support request but we need SOS help...

We used DBISAM 3.x in 180 places in our country without big problems.

But with 3 months before we have got incomprehensible errors in one of our customer.

This corp have 20-30 active user, Windows 2003 operation systems (in clients and in server too), good servers, star-topography network, and
good qualified system admin. They have Nod32 clients (and none of AV in server), with Windows domain system.

Formerly they are used windows sharing. In that time we got many Index Page Buffer error - because of one client have wrong network card.

Later we changed to "database server".

This was a hope to we get out from these kind of errors.

But: after we fully changed the system, we got errors too...

Sometimes before the AV-s we got Index Page Buffer errors. Sometimes we got nothing, or we got only a record locking errors.

The sysadmin changed the dbsrvr's place from server to an IBM server (hardware RAID 5) to avoid the conflicts with other apps. The bugs/errors
are continued.

Interesting that sysadmin sent me the database two times after Index Page Buffer error, I don't found any errors in the database... Hmmm...

These errors are daily repeating, many records/work are lost (because of server and client AV-s), and they are cannot work.

We tried everything we can, but we are out of ideas... In 179 places this is working (many of these also have Windows 2003 without problems),
but only this customer have this big problem - they can't work...

Please help me with some idea!!! We want to go to this place in this week, and we want to get any info or idea to trying, and to solving these kind
of errors...

Thanks for it!!!

dd


If filtered some info from the log:

Index Page buffer based errors:

a.)

2009.04.15. 8:26:12 Engine error [DBISAM Engine Error # 8965 Index page buffers corrupt in the table 'termel'] for address [192.168.100.65]
using engine version [3,24] - last request code was [REQUEST_FIND]
2009.04.15. 8:26:12 Engine error [Access violation at address 000DCA7D in module 'dbsrvr.exe'. Write of address 0000003C] for address
[192.168.100.65] using engine version [3,24] - last request code was [REQUEST_APPENDRECORD]
2009.04.15. 8:26:12 Engine error [DBISAM Engine Error # 9729 Duplicate key found in the index '' of the table 'termel'] for address
[192.168.100.65] using engine version [3,24] - last request code was [REQUEST_APPENDRECORD]
2009.04.15. 8:26:32 Engine error [Access violation at address 00013140 in module 'dbsrvr.exe'. Read of address FFFFFFFC] for address
[192.168.100.65] using engine version [3,24] - last request code was [REQUEST_CLOSECURSOR]

2009.05.06. 14:54:01 Engine error [DBISAM Engine Error # 8965 Index page buffers corrupt in the table 'termel'] for address [192.168.100.113]
using engine version [3,24] - last request code was [REQUEST_SETRANGE]
2009.05.06. 14:54:13 Engine error [Access violation at address 000DCA7D in module 'dbsrvr.exe'. Write of address 0000003C] for address
[192.168.100.113] using engine version [3,24] - last request code was [REQUEST_FIND]
2009.05.06. 14:54:17 Engine error [Access violation at address 00013140 in module 'dbsrvr.exe'. Read of address FFFFFFFC] for address
[192.168.100.113] using engine version [3,24] - last request code was [REQUEST_CLOSECURSOR]
2009.05.06. 14:54:17 Engine error [Access violation at address 00013140 in module 'dbsrvr.exe'. Read of address FFFFFFFC] for address
[192.168.100.113] using engine version [3,24] - last request code was [REQUEST_FREECURSOR]
2009.05.06. 14:54:17 Engine error [Access violation at address 00013140 in module 'dbsrvr.exe'. Read of address FFFFFFFC] for address
[192.168.100.113] using engine version [3,24] - last request code was [REQUEST_CLOSEDATADIR]
2009.05.06. 14:54:17 Internal error removing session [Access violation at address 00013140 in module 'dbsrvr.exe'. Read of address FFFFFFFC]
last connected as user [scroll] at address [192.168.100.113]

2009.05.07. 8:36:44 Engine error [DBISAM Engine Error # 8965 Index page buffers corrupt in the table 'termel'] for address [192.168.100.113]
using engine version [3,24] - last request code was [REQUEST_FIND]
2009.05.07. 8:36:44 Engine error [Access violation at address 000D93A8 in module 'dbsrvr.exe'. Read of address 00000118] for address
[192.168.100.113] using engine version [3,24] - last request code was [REQUEST_ACTIVATEFILTER]
2009.05.07. 8:36:47 Engine error [Access violation at address 00013140 in module 'dbsrvr.exe'. Read of address FFFFFFFC] for address
[192.168.100.113] using engine version [3,24] - last request code was [REQUEST_CLOSECURSOR]
2009.05.07. 8:36:47 Engine error [Access violation at address 00013140 in module 'dbsrvr.exe'. Read of address FFFFFFFC] for address
[192.168.100.113] using engine version [3,24] - last request code was [REQUEST_FREECURSOR]
2009.05.07. 8:36:47 Engine error [Access violation at address 00013140 in module 'dbsrvr.exe'. Read of address FFFFFFFC] for address
[192.168.100.113] using engine version [3,24] - last request code was [REQUEST_CLOSEDATADIR]
2009.05.07. 8:36:47 Internal error removing session [Access violation at address 00013140 in module 'dbsrvr.exe'. Read of address FFFFFFFC]
last connected as user [scroll] at address [192.168.100.113]

2009.05.05. 10:09:48 Engine error [DBISAM Engine Error # 8965 Index page buffers corrupt in the table 'termel'] for address [192.168.100.65]
using engine version [3,24] - last request code was [REQUEST_FIND]
2009.05.05. 10:09:48 Engine error [DBISAM Engine Error # 8965 Index page buffers corrupt in the table 'termel'] for address [192.168.100.65]
using engine version [3,24] - last request code was [REQUEST_ADDEXPRFILTER]
2009.05.05. 10:09:51 Engine error [Access violation at address 00013140 in module 'dbsrvr.exe'. Read of address FFFFFFFC] for address
[192.168.100.65] using engine version [3,24] - last request code was [REQUEST_CLOSECURSOR]
2009.05.05. 10:09:51 Engine error [Access violation at address 00013140 in module 'dbsrvr.exe'. Read of address FFFFFFFC] for address
[192.168.100.65] using engine version [3,24] - last request code was [REQUEST_FREECURSOR]
2009.05.05. 10:09:51 Engine error [Access violation at address 00013140 in module 'dbsrvr.exe'. Read of address FFFFFFFC] for address
[192.168.100.65] using engine version [3,24] - last request code was [REQUEST_CLOSEDATADIR]


b.) Unknown errors:
2009.05.08. 12:18:36 Engine error [Access violation at address 000122BC in module 'dbsrvr.exe'. Write of address 00000000] for address
[192.168.100.53] using engine version [3,24] - last request code was [REQUEST_OPENCURSOR]
2009.05.08. 12:18:39 Engine error [Invalid pointer operation] for address [192.168.100.53] using engine version [3,24] - last request code was
[REQUEST_FREESTMT]
2009.05.08. 12:18:39 Engine error [Invalid pointer operation] for address [192.168.100.53] using engine version [3,24] - last request code was
[REQUEST_CLOSECURSOR]
2009.05.08. 12:18:39 Engine error [Invalid pointer operation] for address [192.168.100.53] using engine version [3,24] - last request code was
[REQUEST_FREECURSOR]

2009.05.05. 12:53:03 Engine error [DBISAM Engine Error # 10258 Record in the table 'adm' is in use by another user] for address
[192.168.100.52] using engine version [3,24] - last request code was [REQUEST_GETCURRECORD]
2009.05.05. 13:04:19 Engine error [Access violation at address 000D3EB9 in module 'dbsrvr.exe'. Write of address 0000001C] for address
[192.168.100.65] using engine version [3,24] - last request code was [REQUEST_FIND]

2009.04.17. 7:19:15 Engine error [DBISAM Engine Error # 10258 Record in the table 'adm' is in use by another user] for address
[192.168.100.85] using engine version [3,24] - last request code was [REQUEST_GETCURRECORD]
2009.04.17. 7:39:57 Engine error [Access violation at address 0010682F in module 'dbsrvr.exe'. Read of address 09B27D7C] for address
[192.168.100.65] using engine version [3,24] - last request code was [REQUEST_FIND]
2009.04.17. 7:40:04 Engine error [Access violation at address 000D32A7 in module 'dbsrvr.exe'. Write of address 00000001] for address
[192.168.100.65] using engine version [3,24] - last request code was [REQUEST_GETNEXTRECORD]
2009.04.17. 7:40:08 Engine error [Invalid pointer operation] for address [192.168.100.65] using engine version [3,24] - last request code was
[REQUEST_CLOSECURSOR]
2009.04.17. 7:40:08 Engine error [Invalid pointer operation] for address [192.168.100.65] using engine version [3,24] - last request code was
[REQUEST_FREECURSOR]
2009.04.17. 7:40:08 Engine error [Invalid pointer operation] for address [192.168.100.65] using engine version [3,24] - last request code was
[REQUEST_CLOSEDATADIR]

2009.04.16. 8:24:54 Engine error [DBISAM Engine Error # 10258 Record in the table 'adm' is in use by another user] for address
[192.168.100.85] using engine version [3,24] - last request code was [REQUEST_GETCURRECORD]
2009.04.16. 8:27:45 Engine error [Access violation at address 000D32A7 in module 'dbsrvr.exe'. Write of address 00000001] for address
[192.168.100.85] using engine version [3,24] - last request code was [REQUEST_SETTOBOOKMARK]
2009.04.16. 8:35:05 Engine error [DBISAM Engine Error # 10258 Record in the table 'adm' is in use by another user] for address
[192.168.100.79] using engine version [3,24] - last request code was [REQUEST_GETCURRECORD]

durumdara


If its working well in 179 places and badly in 1 place that suggests the problem is with the 1 place rather than the application or its components. I'd suspect something different in their configuration / usage.

Sometimes its simply volume that causes the problem, other times the OS and its supporting software (Anti Virus etc), others its the hardware.

As a start point I'd try and identify what the differences are between the sites that are OK and this rouge site.

The other thing that's always worth trying is to upgrade to the latest V3 (I think its 3.30) and see if that makes a difference.

Roy Lambert [Team Elevate]

>
>Roy Lambert wrote:
>
>If its working well in 179 places and badly in 1 place that suggests the problem is with the 1 place rather than the application or its components.
>I'd suspect something different in their configuration / usage.

Usage is same, application is same. Server.exe, config is same. But all of other things are different.

>Sometimes its simply volume that causes the problem, other times the OS and its supporting software (Anti Virus etc), others its the hardware.
>As a start point I'd try and identify what the differences are between the sites that are OK and this rouge site.

This is very hard thing. The comparation of two different network is impossible. Also impossible to stop the network, or the stop the work in these
places... The customer is print corporation, and "show must go on!"...

Is it possible that problem source is not the hardware - this is an OS problem...?
Possible something changed in some SP, and dbsrvr not get that it wait for???

Do you heard about OS conflicts?

Is it possible thing if a client have wrong hardware (network card or other) that can "damage" or "disturb" the server service?

Which are the things that can injure dbsrvr or can damage it's inner data structure/or tables? Can I confuse DBSRVR in software way?

Do you heard about NOD32 conflict (in clients only)?

Thanks for your help: dd

durumdara


>Usage is same, application is same. Server.exe, config is same. But all of other things are different.

Which does rather suggest its their site.

>>Sometimes its simply volume that causes the problem, other times the OS and its supporting software (Anti Virus etc), others its the hardware.
>>As a start point I'd try and identify what the differences are between the sites that are OK and this rouge site.
>
>This is very hard thing. The comparation of two different network is impossible. Also impossible to stop the network, or the stop the work in these
>places... The customer is print corporation, and "show must go on!"...

Its hard, but its a very good start point

>Is it possible that problem source is not the hardware - this is an OS problem...?
>Possible something changed in some SP, and dbsrvr not get that it wait for???

Its possible but less likely than hardware. Its also possible, but again less likely, that its just some weird thing they're doing that isn't happening at any other site and that's messing things up. I had a situation where I had background threads running, if I had four threads running then under some circumstances I had a page index error. Took months to identify the culprit although fixing it was pretty fast.

>Do you heard about OS conflicts?

No

>Is it possible thing if a client have wrong hardware (network card or other) that can "damage" or "disturb" the server service?

I seem to recall a post from Tim about this not to long ago and I think the answer was possibly. DBISAM can only work with what its given, but since everything's happening on the server it should be consistent.

I've also heard of the hard drive throwing up these sort of problems. Usually just before it dies.

I also have a vague memory of some other software "stomping" on memory belonging to other programs which could cause weird errors.

>Which are the things that can injure dbsrvr or can damage it's inner data structure/or tables? Can I confuse DBSRVR in software way?

I wouldn't have thought so.

>Do you heard about NOD32 conflict (in clients only)?

Yes there have been some, it mainly seems to be access to temporary tables being prevented - search these ng's

Sorry I'm not being very helpful. Why not email Tim direct and see if he can come up with something?

Roy Lambert [Team Elevate]

Roy Lambert wrote:
> durumdara
>
> ...
>> Do you heard about NOD32 conflict (in clients only)?
>
> Yes there have been some, it mainly seems to be access to temporary tables being prevented - search these ng's
>

I seem to remember NOD32 interfering with TCP/IP traffic, but may be
wrong (or irrelevant).  Excluding AntiVirus-scan of any folder with
database tables (temp or perm) is usually a smart thing.


--
Aage J.

If you are having issues like this and have an active support plan, you
should always contact support@elevatesoft.com *first* before posting here.
That way we can work on a solution without you having to wait any longer
than necessary.

I've got some new debugging tools that I've been working on for DBISAM 4.x
as part of a new effort to mop up any weird issues like this, and I can port
the tools back to DBISAM 3.x also.  Contact me via email and I will work
with you on getting you a debug version of the 3.x DBISAM Database Server
that will at least get us some additional information (call stack, etc.)
that we can use to figure out where the issue is.

--
Tim Young
Elevate Software
www.elevatesoft.com