Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

DB2 Express-C 9.7.5 Urgent Problem ECF_FILE_DOESNT_EXIST

1,371 views
Skip to first unread message

Fin

unread,
Dec 3, 2011, 10:19:04 AM12/3/11
to
Hi folks, urgent problem here.

I just yesterday upgraded from Express-C 32-bit on XP SP3 9.7.4 to 9.7.5 and all of a sudden I have errors pouring out of the db2diag.log file apparently complaining about the tablespace locations. Errors listed below.

Further, the database still appears to be working however, just cannot get rid of the error. I just installed Express-C 64-bit 9.7.5 on another machine with Windows 7 Pro, SP1 and that too is having the same problems.

My other issue is the I backed up my database on the XP machine so I could perform a redirected restore to the new 64-bit, which kind of worked but now it says that I need to drop and recreate every single stored procedure as it fails with an SQL0471N error saying "The invocation failed because the SQL routine was created on a previous release of the product and the DBMS was not able to migrate it to the current release. The routine must be dropped and re-created to become operative again."

But I upgraded to 9.7.5 before I backed it up and moved it to the new machine ?

The XP machine SPs all work just fine from what I can tell, but the restored SPs from the 9.7.5 32-bit image do not. Any ideas ? I cannot face having to recreate a few hundred stored procedures !

Many thanks for any help. Fin.
==================================
2011-12-03-08.44.53.044000-360 I24459F1315 LEVEL: Error (OS)
PID : 2288 TID : 4124 PROC : db2syscs.exe
INSTANCE: DB2 NODE : 000 DB : SQ4TEST
APPHDL : 0-7 APPID: *LOCAL.DB2.111203144449
AUTHID : DB2ADMIN
EDUID : 4124 EDUNAME: db2agent (SQ4TEST)
FUNCTION: DB2 Common, OSSe, ossPathTest, probe:40
MESSAGE : ECF=0x90000007=-1879048185=ECF_UNKNOWN
Unknown
CALLED : OS, -, access
OSERR : 123 "The filename, directory name, or volume label syntax is incorrect."
DATA #1 : String, 45 bytes
C:\DB2\CONTAINERS\SQ4TEST\TABLESPACES\STATIC\
DATA #2 : Hex integer, 8 bytes
0x0000000000000001
DATA #3 : Hex integer, 8 bytes
0x0000000000000000
DATA #4 : signed integer, 4 bytes
0
CALLSTCK:
[0] 0x00000001801269CF pdOSSeLoggingCallback + 0x265
[1] 0x000000000054DC99 ossLogSysRC + 0xF3
[2] 0x000000000054BBA1 ossSystemErrorHandler + 0x1D7
[3] 0x0000000000528578 ossPathTest + 0x226
[4] 0x00000000005282DE ossPathExists + 0x3C
[5] 0x00000000005380BC ossGetDiskInfo + 0x3EE
[6] 0x000000018003F110 sqloFetchAndStoreFSInfoInFileHandle + 0x134
[7] 0x0000000001CCE41D sqlbinit + 0x4D05
[8] 0x0000000001D84F5D sqlbpini + 0x28EC7
[9] 0x0000000001CF2224 sqlbFreeOnePool + 0x7D8

2011-12-03-08.44.53.059000-360 I25776F597 LEVEL: Severe
PID : 2288 TID : 4124 PROC : db2syscs.exe
INSTANCE: DB2 NODE : 000 DB : SQ4TEST
APPHDL : 0-7 APPID: *LOCAL.DB2.111203144449
AUTHID : DB2ADMIN
EDUID : 4124 EDUNAME: db2agent (SQ4TEST)
FUNCTION: DB2 UDB, buffer pool services, sqlbFetchAndStoreContainerFSInfo, probe:10
MESSAGE : ECF=0x9000001A=-1879048166=ECF_FILE_DOESNT_EXIST
File doesn't exist
DATA #1 : File name, 44 bytes
C:\DB2\CONTAINERS\SQ4TEST\TABLESPACES\STATIC

2011-12-03-08.44.53.075000-360 I26375F1289 LEVEL: Error (OS)
PID : 2288 TID : 4124 PROC : db2syscs.exe
INSTANCE: DB2 NODE : 000 DB : SQ4TEST
APPHDL : 0-7 APPID: *LOCAL.DB2.111203144449
AUTHID : DB2ADMIN
EDUID : 4124 EDUNAME: db2agent (SQ4TEST)
FUNCTION: DB2 Common, OSSe, ossPathTest, probe:40
MESSAGE : ECF=0x90000007=-1879048185=ECF_UNKNOWN
Unknown
CALLED : OS, -, access
OSERR : 123 "The filename, directory name, or volume label syntax is incorrect."
DATA #1 : String, 45 bytes
C:\DB2\CONTAINERS\SQ4TEST\TABLESPACES\STATIC\
DATA #2 : Hex integer, 8 bytes
0x0000000000000001
DATA #3 : Hex integer, 8 bytes
0x0000000000000000
DATA #4 : signed integer, 4 bytes
0
CALLSTCK:
[0] 0x00000001801269CF pdOSSeLoggingCallback + 0x265
[1] 0x000000000054DC99 ossLogSysRC + 0xF3
[2] 0x000000000054BBA1 ossSystemErrorHandler + 0x1D7
[3] 0x0000000000528578 ossPathTest + 0x226
[4] 0x00000000005282DE ossPathExists + 0x3C
[5] 0x00000000005380BC ossGetDiskInfo + 0x3EE
[6] 0x0000000001D7312E sqlbpini + 0x17098
[7] 0x0000000001D8524D sqlbpini + 0x291B7
[8] 0x0000000001CF2224 sqlbFreeOnePool + 0x7D8
[9] 0x0000000001CCA403 sqlbinit + 0xCEB

2011-12-03-08.44.53.075000-360 I27666F745 LEVEL: Severe
PID : 2288 TID : 4124 PROC : db2syscs.exe
INSTANCE: DB2 NODE : 000 DB : SQ4TEST
APPHDL : 0-7 APPID: *LOCAL.DB2.111203144449
AUTHID : DB2ADMIN
EDUID : 4124 EDUNAME: db2agent (SQ4TEST)
FUNCTION: DB2 UDB, buffer pool services, sqlbCountContsOnCriticalPaths, probe:30
MESSAGE : Could not get fsID info for container
DATA #1 : File name, 44 bytes
C:\DB2\CONTAINERS\SQ4TEST\TABLESPACES\STATIC
DATA #2 : unsigned integer, 4 bytes
2415919130
DATA #3 : unsigned integer, 4 bytes
2415919106
DATA #4 : Pool ID, PD_TYPE_SQLB_POOL_ID, 2 bytes
34
DATA #5 : unsigned integer, 4 bytes
0
=======================

Mark A

unread,
Dec 3, 2011, 4:58:15 PM12/3/11
to
I would recommend that you contact the DB2 Express-C forum on the IBM website. Also, make sure you
read the installation directions carefully, because in the past the install/upgrade directions for Express-
C may be different than for other DB2 editions.

Fin

unread,
Dec 3, 2011, 6:53:23 PM12/3/11
to
Thanks Mark, already posted over there. Read the instructions from cover to cover and it said that as long as you were upgrading within release, ie: 9.7.4 to 9.7.5 then it should work fine.

Anyway, I had to re-create ALL my 300+ stored procedures which wasn't pretty.

The ADMIN server is completely screwed up as far as I can tell, the health monitor is producing errors all over the place as well. Seems related to me. Thousands of these messages.

2011-12-03-17.50.39.227000-360 I25956F432 LEVEL: Warning
PID : 3288 TID : 4732 PROC : db2fmp64.exe
INSTANCE: DB2 NODE : 000
EDUID : 4732
FUNCTION: DB2 UDB, Health Monitor, HmonMainCB::refreshDbAutonomicSwitches, probe:160
MESSAGE : Failed connecting to database "SQ4TEST "
DATA #1 : Hexdump, 4 bytes
0x000000000731D6EC : 7E8A FFFF ~...

Fin

unread,
Dec 4, 2011, 9:26:09 AM12/4/11
to
Problem solved with the ECF error, the DBDAS service with installed, owned and started by a local service, needed to be db2admin to have the right access rights to the table spaces.

Above issue still remains with the Health Monitor.

Fin

unread,
Dec 4, 2011, 9:55:47 AM12/4/11
to
Grrrrrr .. problem is back again. I don't get this, and OS error saying the file does not exist when the database itself is working fine. Am I missing a path or something in the db2 registry ?

I thought about trying to rebind but I can't get the bind option to work from the Java Configuration Assistant either. Just sits there with the hourglass.

2011-12-04-08.46.40.976000-360 I20866F1321 LEVEL: Error (OS)
PID : 4712 TID : 3664 PROC : db2syscs.exe
INSTANCE: DB2 NODE : 000 DB : SQ4TEST
APPHDL : 0-153 APPID: *LOCAL.DB2.111204144639
AUTHID : DB2ADMIN
EDUID : 3664 EDUNAME: db2agent (SQ4TEST)
FUNCTION: DB2 Common, OSSe, ossPathTest, probe:40
MESSAGE : ECF=0x90000007=-1879048185=ECF_UNKNOWN
Unknown
CALLED : OS, -, access
OSERR : 123 "The filename, directory name, or volume label syntax is incorrect."
DATA #1 : String, 51 bytes
C:\db2\containers\sq4test\tablespaces\INDEXES_LONG\
DATA #2 : Hex integer, 8 bytes
0x0000000000000001
DATA #3 : Hex integer, 8 bytes
0x0000000000000000
DATA #4 : signed integer, 4 bytes
0
CALLSTCK:
[0] 0x00000001801269CF pdOSSeLoggingCallback + 0x265
[1] 0x000000000045DC99 ossLogSysRC + 0xF3
[2] 0x000000000045BBA1 ossSystemErrorHandler + 0x1D7
[3] 0x0000000000438578 ossPathTest + 0x226
[4] 0x00000000004382DE ossPathExists + 0x3C
[5] 0x00000000004480BC ossGetDiskInfo + 0x3EE
[6] 0x000000018003F110 sqloFetchAndStoreFSInfoInFileHandle + 0x134
[7] 0x0000000001CCE41D sqlbinit + 0x4D05
[8] 0x0000000001D84F5D sqlbpini + 0x28EC7
[9] 0x0000000001CF2224 sqlbFreeOnePool + 0x7D8

2011-12-04-08.46.40.978000-360 I22189F603 LEVEL: Severe
PID : 4712 TID : 3664 PROC : db2syscs.exe
INSTANCE: DB2 NODE : 000 DB : SQ4TEST
APPHDL : 0-153 APPID: *LOCAL.DB2.111204144639
AUTHID : DB2ADMIN
EDUID : 3664 EDUNAME: db2agent (SQ4TEST)
FUNCTION: DB2 UDB, buffer pool services, sqlbFetchAndStoreContainerFSInfo, probe:10
MESSAGE : ECF=0x9000001A=-1879048166=ECF_FILE_DOESNT_EXIST
File doesn't exist
DATA #1 : File name, 50 bytes
C:\db2\containers\sq4test\tablespaces\INDEXES_LONG

Helmut Tessarek

unread,
Dec 10, 2011, 5:59:51 PM12/10/11
to
Posted some question in the other thread, but I'll add them here as well as
this is actually the correct thread:

I will discuss this with DB2 kernel development and hope that I will have an
answer soon.
Just to clarify some things:

The ECF problem occurs on both systems, 32bit and 64bit? Right?
Both databases have been upgraded from a previous version, correct? If yes,
which version/release?

What happens, if you create a new test database on 9.7.5? Do you see the same
error messages?

With regard to the Health Monitor problem, I sadly have to say that it is
deprecated as well and therefore I'm not really sure how to address this.
Furthermore please note that all Control Center Tools are deprecated.

Deprecated functionality - http://j.mp/ub2rIn

--
Helmut K. C. Tessarek
DB2 Performance and Development
IBM Toronto Lab

Fin

unread,
Dec 12, 2011, 4:00:14 AM12/12/11
to
Thanks for the response Helmut and it is great news to hear there is progress.

Yes, the identical ECF issues are occurring on both 32 and 64 bit systems.

The upgrade issue is a little trickier. the 32 bit system was upgraded from a clean Exrpress-C 9.7.4 version, backed up and redirected restore to the the new 64-bit system, supposedly thus inheriting the same ECF problem I would assume. The 32 bit system is XP Pro SP3, 64 is Win 7 Pro SP1.

Unfortunately I did not realize the issue til after the restore.

Created a new DB called TEST which worked fine, but again got the same ECF Issues. See below.

I understand that Control Center was being deprecated, but did not realize the health center, monitor and adviser were also going away. That being the case, how do I get rid of the error messages and stop it ? What if anything is replacing the automatic maintenance options and why is that not working on the 64-bit system ?

Many grateful thanks for your help on this, greatly appreciate it. Fin.

2011-12-12-03.22.36.049000-300 I49747F1307 LEVEL: Error (OS)
PID : 2872 TID : 4264 PROC : db2syscs.exe
INSTANCE: DB2 NODE : 000 DB : TEST
APPHDL : 0-550 APPID: *LOCAL.DB2.111212082235
AUTHID : TIM
EDUID : 4264 EDUNAME: db2agent (TEST)
FUNCTION: DB2 Common, OSSe, ossPathTest, probe:40
MESSAGE : ECF=0x90000007=-1879048185=ECF_UNKNOWN
Unknown
CALLED : OS, -, access
OSERR : 123 "The filename, directory name, or volume label syntax is incorrect."
DATA #1 : String, 43 bytes
C:\DB2\NODE0000\TEST\T0000002\C0000000.LRG\
DATA #2 : Hex integer, 8 bytes
0x0000000000000001
DATA #3 : Hex integer, 8 bytes
0x0000000000000000
DATA #4 : signed integer, 4 bytes
0
CALLSTCK:
[0] 0x00000001801269CF pdOSSeLoggingCallback + 0x265
[1] 0x000000000045DC99 ossLogSysRC + 0xF3
[2] 0x000000000045BBA1 ossSystemErrorHandler + 0x1D7
[3] 0x0000000000438578 ossPathTest + 0x226
[4] 0x00000000004382DE ossPathExists + 0x3C
[5] 0x00000000004480BC ossGetDiskInfo + 0x3EE
[6] 0x000000018003F110 sqloFetchAndStoreFSInfoInFileHandle + 0x134
[7] 0x0000000001CCE41D sqlbinit + 0x4D05
[8] 0x0000000001D84F5D sqlbpini + 0x28EC7
[9] 0x0000000001CF2224 sqlbFreeOnePool + 0x7D8

2011-12-12-03.22.36.051000-300 I51056F589 LEVEL: Severe
PID : 2872 TID : 4264 PROC : db2syscs.exe
INSTANCE: DB2 NODE : 000 DB : TEST
APPHDL : 0-550 APPID: *LOCAL.DB2.111212082235
AUTHID : TIM
EDUID : 4264 EDUNAME: db2agent (TEST)
FUNCTION: DB2 UDB, buffer pool services, sqlbFetchAndStoreContainerFSInfo, probe:10
MESSAGE : ECF=0x9000001A=-1879048166=ECF_FILE_DOESNT_EXIST
File doesn't exist
DATA #1 : File name, 42 bytes
C:\DB2\NODE0000\TEST\T0000002\C0000000.LRG


Fin

unread,
Dec 12, 2011, 3:20:31 PM12/12/11
to
Helmut,

Just so you know, I performed a full UN-install, reboot and clean install of 9.7.5 on the 64-bit system. Created the SAMPLE database from First Steps and that too has the exact same ECF issues. Beginning to think this is not just me and that 9.7.5 has some serious problems.

Any way you can point to me downloading the install files for 9.7.4 both 32 and 64-bit ?

Helmut Tessarek

unread,
Dec 13, 2011, 3:10:28 PM12/13/11
to
On 12.12.11 15:20 , Fin wrote:
> Any way you can point to me downloading the install files for 9.7.4 both 32 and 64-bit ?

Working on it...

Helmut Tessarek

unread,
Dec 13, 2011, 3:39:50 PM12/13/11
to
On 12.12.11 4:00 , Fin wrote:
> Yes, the identical ECF issues are occurring on both 32 and 64 bit systems.

The ECF error messages seem to be a know bug and there is an APAR IC79739.
http://www-01.ibm.com/support/docview.wss?uid=swg1IC79739

You don't have to worry. The messages are not very nice, but your data is not
at risk. The only way to get rid of these messages right now is to downgrade
to 9.7.4.
I'll let you know as soon as I have the location of the old images.

> I understand that Control Center was being deprecated, but did not realize
> the health center, monitor and adviser were also going away. That being the
> case, how do I get rid of the error messages and stop it ? What if anything
> is replacing the automatic maintenance options and why is that not working
> on the 64-bit system ?

Automatic maintenance will not be replaced, at least I am not aware of it.
Automatic maintenance does not require any of the deprecated functionality, so
everything is fine.

To get rid of any health monitor related messages just turn off the health
monitor in the database manager configuration.

Fin

unread,
Dec 13, 2011, 4:07:19 PM12/13/11
to
Thanks for the response Helmut, very much appreciated.

That's good news that it is now a known issue, I was beginning to think it was just me. Funnily enough after I did the re-install yesterday, automatic maintenance has all of sudden started working again. Hit my window the afternoon and away it went, the health monitor issues have also gone away, leaving us with the ECF problems.

Thanks for your continued help, regards Fin.

Fin

unread,
Dec 13, 2011, 4:39:23 PM12/13/11
to
There seems to be to other issues with 9.7.5 as well, marking the database BAD and completely crashing. Sooner we can get back to 9.7.4 the better in my book.

Issue had to do with db2_compatibility_vector in the registry.
0 new messages