Urgent Help Connect refused with SQL0290N

Fabrizio Napolitano

Urgent Help Connect refused with SQL0290N
Hello,

I'm trying to solve a couple of problems that are giving me a bit of head
ache, so any advice could be much appreciated...

In both cases the platform is AIX 4.3.3 with DB2 UDB EE v7.1 fp 3

The first and the very important and urgent one is this:

The system crashed and now there are some tablespaces left in a pending
state . No connection are allowed with code SQL0290N and for some "strange"
reason the db2diag.log permit was set to root and so no log was written
since 23 of october.
Now at the moment I can't access the system (I'm working as telephone
support) in any case the people on site assured me that they didn't get any
problem with the disk subsystem and that all the containers are accessibles
and with the right permissions!
My First question if something got wrong on some tablespaces shouldn't be
allowed the connection to the db and only those tablespaces offline??
Second and obvious question there is any way to be able to connect to the db
and so get the list of tablespaces offline or in load or restore pending
state and trying to fix them???
Not knowing wich tbs is troubled I told them to run a db2dart in Inspect
mode to get the situation. But then??? The only action that I can think of
is a restart db putting the damaged tablespaces in drop pending state, is
there any less painfull action????
Is this possible to use the db2dart for a repair action or should I push
them to call IBM Support??

Second question I don't think connected to the first but I don't know...
Starting the istance I get the error:
SQL5043N Support for one or more communications protocols failed to start
successfully. However, core database manager functionality started
successfully.
Well I tried stopping again the istance , running a ps and killing all the
defunct or hung db2 process, ipcs and again after everything is clean and
nice at the next start I got again the same error... where to start to llok
for a clue??
Is it possible that there are contentions with DAS or other instances???

Thanks in advance to all of you
Fabrizio



Jeremy Schleicher

Re: Urgent Help Connect refused with SQL0290N
(in response to Fabrizio Napolitano)
Db2dart can fix the bad tablespace if it can find it. However, I would
have them call IBM support, if the diag log is useless then there is no way
to know which table space it bad, unless you want to do a redirected
restore and get a list that way, that is if you have a good backup. I have
seen a few times that db2dart will not find that object that is damaged. I
had to use the diag log to identify the bad object and it only listed the
object id. Are there any trap files that you can use to identify the
object in question?



From: "Napolitano, Fabrizio" <[login to unmask email]>@RYCI.COM on
12/17/2001 10:40 AM

Please respond to DB2 Data Base Discussion List <[login to unmask email]>

Sent by: DB2 Data Base Discussion List <[login to unmask email]>


To: [login to unmask email]
cc:

Subject: Urgent Help Connect refused with SQL0290N


Hello,

I'm trying to solve a couple of problems that are giving me a bit of head
ache, so any advice could be much appreciated...

In both cases the platform is AIX 4.3.3 with DB2 UDB EE v7.1 fp 3

The first and the very important and urgent one is this:

The system crashed and now there are some tablespaces left in a pending
state . No connection are allowed with code SQL0290N and for some "strange"
reason the db2diag.log permit was set to root and so no log was written
since 23 of october.
Now at the moment I can't access the system (I'm working as telephone
support) in any case the people on site assured me that they didn't get any
problem with the disk subsystem and that all the containers are accessibles
and with the right permissions!
My First question if something got wrong on some tablespaces shouldn't be
allowed the connection to the db and only those tablespaces offline??
Second and obvious question there is any way to be able to connect to the
db
and so get the list of tablespaces offline or in load or restore pending
state and trying to fix them???
Not knowing wich tbs is troubled I told them to run a db2dart in Inspect
mode to get the situation. But then??? The only action that I can think of
is a restart db putting the damaged tablespaces in drop pending state, is
there any less painfull action????
Is this possible to use the db2dart for a repair action or should I push
them to call IBM Support??

Second question I don't think connected to the first but I don't know...
Starting the istance I get the error:
SQL5043N Support for one or more communications protocols failed to start
successfully. However, core database manager functionality started
successfully.
Well I tried stopping again the istance , running a ps and killing all the
defunct or hung db2 process, ipcs and again after everything is clean and
nice at the next start I got again the same error... where to start to llok
for a clue??
Is it possible that there are contentions with DAS or other instances???

Thanks in advance to all of you
Fabrizio








Fabrizio Napolitano

Re: Urgent Help Connect refused with SQL0290N
(in response to Jeremy Schleicher)
Thanks for the reply,

Yes I wouldn't like to try to use db2dart in repair mode either, in that
case I agree I'll make them call IBM support...

The db2dart gave us only a warning that one tbs was demaged... we tried to
restart db drop pending that tbs but it failed again.
Luckly I told the people on site that the first priority was to give
db2diag.log back the right permissions... In this way even if the restart
failed again I got logged all the other damaged tbs!

Considering that it's a test DB and there are only 3 tablespaces (just
discovered it now) are demaged I think we'll get out of this faster dropping
those tbs and reloading the tables afterwards!!
Yes, I can confirm that the restart db with all the damaged tbs brought the
db back to life again!!!

Only one thought still trouble me... the troubleshooting guide says that
from v6 on a damaged tbs shouldn't stop connecting to the db but
"automagically" put those tbs offline... can anyone help me understanding
why it's not the situation??

Thanks again
Fabrizio


-----Original Message-----
From: Jeremy Schleicher
To: [login to unmask email]
Sent: 17/12/01 18.11
Subject: Re: Urgent Help Connect refused with SQL0290N

Db2dart can fix the bad tablespace if it can find it. However, I would
have them call IBM support, if the diag log is useless then there is no
way
to know which table space it bad, unless you want to do a redirected
restore and get a list that way, that is if you have a good backup. I
have
seen a few times that db2dart will not find that object that is damaged.
I
had to use the diag log to identify the bad object and it only listed
the
object id. Are there any trap files that you can use to identify the
object in question?



From: "Napolitano, Fabrizio" <[login to unmask email]>@RYCI.COM on
12/17/2001 10:40 AM

Please respond to DB2 Data Base Discussion List <[login to unmask email]>

Sent by: DB2 Data Base Discussion List <[login to unmask email]>


To: [login to unmask email]
cc:

Subject: Urgent Help Connect refused with SQL0290N


Hello,

I'm trying to solve a couple of problems that are giving me a bit of
head
ache, so any advice could be much appreciated...

In both cases the platform is AIX 4.3.3 with DB2 UDB EE v7.1 fp 3

The first and the very important and urgent one is this:

The system crashed and now there are some tablespaces left in a pending
state . No connection are allowed with code SQL0290N and for some
"strange"
reason the db2diag.log permit was set to root and so no log was written
since 23 of october.
Now at the moment I can't access the system (I'm working as telephone
support) in any case the people on site assured me that they didn't get
any
problem with the disk subsystem and that all the containers are
accessibles
and with the right permissions!
My First question if something got wrong on some tablespaces shouldn't
be
allowed the connection to the db and only those tablespaces offline??
Second and obvious question there is any way to be able to connect to
the
db
and so get the list of tablespaces offline or in load or restore pending
state and trying to fix them???
Not knowing wich tbs is troubled I told them to run a db2dart in Inspect
mode to get the situation. But then??? The only action that I can think
of
is a restart db putting the damaged tablespaces in drop pending state,
is
there any less painfull action????
Is this possible to use the db2dart for a repair action or should I push
them to call IBM Support??

Second question I don't think connected to the first but I don't know...
Starting the istance I get the error:
SQL5043N Support for one or more communications protocols failed to
start
successfully. However, core database manager functionality started
successfully.
Well I tried stopping again the istance , running a ps and killing all
the
defunct or hung db2 process, ipcs and again after everything is clean
and
nice at the next start I got again the same error... where to start to
llok
for a clue??
Is it possible that there are contentions with DAS or other instances???

Thanks in advance to all of you
Fabrizio



the DB2-L webpage at http://www.ryci.com/db2-l. The owners of the list
can




the DB2-L webpage at http://www.ryci.com/db2-l. The owners of the list
can



Fabrizio Napolitano

Re: Urgent Help Connect refused with SQL0290N
(in response to Fabrizio Napolitano)
Thanks for the reply,

Yes I wouldn't like to try to use db2dart in repair mode either, in that
case I agree I'll make them call IBM support...

The db2dart gave us only a warning that one tbs was demaged... we tried to
restart db drop pending that tbs but it failed again.
Luckly I told the people on site that the first priority was to give
db2diag.log back the right permissions... In this way even if the restart
failed again I got logged all the other damaged tbs!

Considering that it's a test DB and there are only 3 tablespaces (just
discovered it now) are demaged I think we'll get out of this faster dropping
those tbs and reloading the tables afterwards!!
Yes, I can confirm that the restart db with all the damaged tbs brought the
db back to life again!!!

Only one thought still trouble me... the troubleshooting guide says that
from v6 on a damaged tbs shouldn't stop connecting to the db but
"automagically" put those tbs offline... can anyone help me understanding
why it's not the situation??

Thanks again
Fabrizio


-----Original Message-----
From: Jeremy Schleicher
To: [login to unmask email]
Sent: 17/12/01 18.11
Subject: Re: Urgent Help Connect refused with SQL0290N

Db2dart can fix the bad tablespace if it can find it. However, I would
have them call IBM support, if the diag log is useless then there is no
way
to know which table space it bad, unless you want to do a redirected
restore and get a list that way, that is if you have a good backup. I
have
seen a few times that db2dart will not find that object that is damaged.
I
had to use the diag log to identify the bad object and it only listed
the
object id. Are there any trap files that you can use to identify the
object in question?



From: "Napolitano, Fabrizio" <[login to unmask email]>@RYCI.COM on
12/17/2001 10:40 AM

Please respond to DB2 Data Base Discussion List <[login to unmask email]>

Sent by: DB2 Data Base Discussion List <[login to unmask email]>


To: [login to unmask email]
cc:

Subject: Urgent Help Connect refused with SQL0290N


Hello,

I'm trying to solve a couple of problems that are giving me a bit of
head
ache, so any advice could be much appreciated...

In both cases the platform is AIX 4.3.3 with DB2 UDB EE v7.1 fp 3

The first and the very important and urgent one is this:

The system crashed and now there are some tablespaces left in a pending
state . No connection are allowed with code SQL0290N and for some
"strange"
reason the db2diag.log permit was set to root and so no log was written
since 23 of october.
Now at the moment I can't access the system (I'm working as telephone
support) in any case the people on site assured me that they didn't get
any
problem with the disk subsystem and that all the containers are
accessibles
and with the right permissions!
My First question if something got wrong on some tablespaces shouldn't
be
allowed the connection to the db and only those tablespaces offline??
Second and obvious question there is any way to be able to connect to
the
db
and so get the list of tablespaces offline or in load or restore pending
state and trying to fix them???
Not knowing wich tbs is troubled I told them to run a db2dart in Inspect
mode to get the situation. But then??? The only action that I can think
of
is a restart db putting the damaged tablespaces in drop pending state,
is
there any less painfull action????
Is this possible to use the db2dart for a repair action or should I push
them to call IBM Support??

Second question I don't think connected to the first but I don't know...
Starting the istance I get the error:
SQL5043N Support for one or more communications protocols failed to
start
successfully. However, core database manager functionality started
successfully.
Well I tried stopping again the istance , running a ps and killing all
the
defunct or hung db2 process, ipcs and again after everything is clean
and
nice at the next start I got again the same error... where to start to
llok
for a clue??
Is it possible that there are contentions with DAS or other instances???

Thanks in advance to all of you
Fabrizio



the DB2-L webpage at http://www.ryci.com/db2-l. The owners of the list
can




the DB2-L webpage at http://www.ryci.com/db2-l. The owners of the list
can