onbar Logical restore cannot complete because of e
On Windows/IDS 11.70 with HP Data Protector 9, the poster used two DP backup specifications: a scheduled whole-system (-w) level-0 archive and a log backup driven by ALARMPROGRAM. The archive session also backed up logs 5525/5526 (while ALARMPROGRAM hit error 43395, 'log backup already running'), so those two logs landed in the archive barlist, not the log barlist. Physical restore worked, but onbar -r -l failed with -43258/-43168, unable to fetch log 5525. Advice was to check DP Resource Type settings so the archive job has no 'L' (logs) and the log job has only 'L'. The poster then asked whether dropping -w and separating logs would work; no confirmed resolution is recorded.
Auto-generated by DrWatson from the posts below — may be imperfect; read the full thread.
Hello all,
OS: Windows Server 2008 R2
DB: Informix IDS 11.70 FC8GE
SM: HP Data Protector 9
We have two separate backup specifications - one for full backups, one for
log backups. The instance is configured with environment variable parameter
OB2BARLIST poining to the log specification in the instance .cmd console
file and in ALARMPROGRAM.
The full backups are being run by Scheduler of Data protector each day
using the Full backup specifications. The log backups are being triggered
by ALARMPROGRAM when a log fills up.
During our monthly restore tests, we've hit an error which prevents us from
performing Logical restore of the instance. The problem is that onbar
cannot find two logs from the Storage manager, which are the first two logs
after the time Full backup was taken. The problem is that when Full backup
started, ALARMPROGRAM started trying to backup current log log 5525 (the
current log by the time when Full started). Then, another log 5526 was
filled up and also was backed up by the Full backup session. In the
meantime, ALARMPROGRAM caused error 43395 An logical log backup is already
running - cannot start another.
So at the end we have the Full backup specification holding all the
dbspaces, critical files, and logs 5525 and 5526 and the Log specification
holding up all logical logs besides these.
The Physical restore (onbar -r -p) using the Full backup specification
succeeds without worries, but after when we try logical restore (onbar -r
-l) onbar complains it can't find log 5525:
2015-11-11 15:19:02 180 180 Begin reserved pages restore (level 0 of
rootdbs,
Storage Manager copy ID: 2866 2015011010).
2015-11-11 15:19:03 180 180 Completed reserved pages restore (level 0 of
rootd
bs).
2015-11-11 15:19:04 180 180 Successfully connected to Storage Manager.
2015-11-11 15:19:39 180 180 Begin cold level 0 restore rootdbs (Storage
Manage
r copy ID: 2866 2015011010).
2015-11-11 15:19:59 180 180 Completed cold level 0 restore rootdbs.
2015-11-11 15:20:35 180 180 Begin cold level 0 restore
ona2aqifx030p0dbs01 (St
orage Manager copy ID: 2866 2015011010).
2015-11-11 15:52:02 180 180 Completed cold level 0 restore
ona2aqifx030p0dbs01
.
2015-11-11 15:52:38 180 180 Begin cold level 0 restore sbspace01 (Storage
Mana
ger copy ID: 2866 2015011010).
2015-11-11 15:52:40 180 180 Completed cold level 0 restore sbspace01.
2015-11-11 15:53:16 180 180 Begin cold level 0 restore logdbs01 (Storage
Manag
er copy ID: 2866 2015011010).
2015-11-11 15:53:18 180 180 Completed cold level 0 restore logdbs01.
2015-11-11 15:53:53 180 180 Begin cold level 0 restore physdbs (Storage
Manage
r copy ID: 2866 2015011010).
2015-11-11 15:53:56 180 180 Completed cold level 0 restore physdbs.
2015-11-11 15:53:58 180 180 Successfully connected to Storage Manager.
2015-11-11 15:53:58 180 180 XBSA Error: (BSAGetObject) Backup object does
not
exist in Storage Manager.
2015-11-11 15:53:58 180 180 Object information: bar_objdesc
obj_id 0 obj_name '5525' obj_type 'L' act_id 0 act_type 2 act_status 0
act_start '2015-11-11 15:53:58' act_end '2015-11-10 21:39:49'
ins_time 0 rsam_time 0 seal_time 1447187988 prev_seal_time 1447155921
level
0 copyid hi:lo 2866:2015011010 req_act_id 7150
logstream 0 est_pages hi:lo 0:0 first_log 0 chpt_log 5525 last_log 0
partial_flag 0 do_query 0 ins_sm_id 0 ins_sm_name ''
ins_verify 0 ins_verify_date '' restore order 0:0
objInfo ''
retry 0 in_catalog 1 in_bootfile 0 child_pid 0 child_state 0
bkup_host '' backup_order 0
* 2015-11-11 15:53:58 180 180 (-43258) ON-Bar could not get logical log
5525 fro*
*m the storage manager.*
2015-11-11 15:54:21 180 180 (-43168) ON-Bar suspended the logical restore
on l
og 5525 (expected to
restore to 2147483647).
2015-11-11 15:54:27 180 180 C:\\\\PROGRA~1\\\\IBM\\\\Informix\\\\11.70\\\\bin\\\\onbar_d
complet
e, returning 100 (0x64)
I tried manually to PITR to log 5525 using either of the specifications(set
OB2BARLIST), but again onbar can't find the logs :(
So my question is: besides my attempts to reschedule the full backup in a
time with lowest workload possible, is there something else I can try to do
to prevent this situation happens in the future?
Or this is limitation of the Data protector integration for Informix?
Thanks in advance for any comments!
--001a114daacc33c074052445ca1e
Hello.
This kind of errors are common to DP, if you are using 2 separate jobs (one
for archives, and the other for logs through alarmprogram).
I just don't understand what job did backup the 2 logs you've mentioned. Did
any onbar process hang in your filesystem?? If so, were they killed manually???
As the error log description says, the next logical log backup should
automatically include the 2 (or more) missed logs. No magic.
Just make sure your archive job in DP does not have a L checked in your
Resource type configuration, and also the opposite (your logical logs job must
only have the L checkbox marked, in its Resource Type configuration).
That should separate and not start logical logs backups by the archive job.
Alexandre Marini
IBM Informix Certified Professional v10 / v11.50 / v11.70 / v12.10
IBM Information Management Informix Technical Professional
IBM Certified Developer - Informix Genero
BRIUG website administrator
Informix independent consultant
> To: ids@iiug.org
> From: lgrigorlu1@gmail.com
> Subject: onbar Logical restore cannot complete because .... [36046]
> Date: Wed, 11 Nov 2015 10:50:51 -0500
>
> Hello all,
>
> OS: Windows Server 2008 R2
> DB: Informix IDS 11.70 FC8GE
> SM: HP Data Protector 9
>
> We have two separate backup specifications - one for full backups, one for
> log backups. The instance is configured with environment variable parameter
> OB2BARLIST poining to the log specification in the instance .cmd console
> file and in ALARMPROGRAM.
> The full backups are being run by Scheduler of Data protector each day
> using the Full backup specifications. The log backups are being triggered
> by ALARMPROGRAM when a log fills up.
>
> During our monthly restore tests, we've hit an error which prevents us from
> performing Logical restore of the instance. The problem is that onbar
> cannot find two logs from the Storage manager, which are the first two logs
> after the time Full backup was taken. The problem is that when Full backup
> started, ALARMPROGRAM started trying to backup current log log 5525 (the
> current log by the time when Full started). Then, another log 5526 was
> filled up and also was backed up by the Full backup session. In the
> meantime, ALARMPROGRAM caused error 43395 An logical log backup is already
> running - cannot start another.
>
> So at the end we have the Full backup specification holding all the
> dbspaces, critical files, and logs 5525 and 5526 and the Log specification
> holding up all logical logs besides these.
>
> The Physical restore (onbar -r -p) using the Full backup specification
> succeeds without worries, but after when we try logical restore (onbar -r
> -l) onbar complains it can't find log 5525:
>
> 2015-11-11 15:19:02 180 180 Begin reserved pages restore (level 0 of
> rootdbs,
> Storage Manager copy ID: 2866 2015011010).
> 2015-11-11 15:19:03 180 180 Completed reserved pages restore (level 0 of
> rootd
> bs).
> 2015-11-11 15:19:04 180 180 Successfully connected to Storage Manager.
> 2015-11-11 15:19:39 180 180 Begin cold level 0 restore rootdbs (Storage
> Manage
> r copy ID: 2866 2015011010).
> 2015-11-11 15:19:59 180 180 Completed cold level 0 restore rootdbs.
> 2015-11-11 15:20:35 180 180 Begin cold level 0 restore
> ona2aqifx030p0dbs01 (St
> orage Manager copy ID: 2866 2015011010).
> 2015-11-11 15:52:02 180 180 Completed cold level 0 restore
> ona2aqifx030p0dbs01
> ..
> 2015-11-11 15:52:38 180 180 Begin cold level 0 restore sbspace01 (Storage
> Mana
> ger copy ID: 2866 2015011010).
> 2015-11-11 15:52:40 180 180 Completed cold level 0 restore sbspace01.
> 2015-11-11 15:53:16 180 180 Begin cold level 0 restore logdbs01 (Storage
> Manag
> er copy ID: 2866 2015011010).
> 2015-11-11 15:53:18 180 180 Completed cold level 0 restore logdbs01.
> 2015-11-11 15:53:53 180 180 Begin cold level 0 restore physdbs (Storage
> Manage
> r copy ID: 2866 2015011010).
> 2015-11-11 15:53:56 180 180 Completed cold level 0 restore physdbs.
> 2015-11-11 15:53:58 180 180 Successfully connected to Storage Manager.
> 2015-11-11 15:53:58 180 180 XBSA Error: (BSAGetObject) Backup object does
> not
> exist in Storage Manager.
> 2015-11-11 15:53:58 180 180 Object information: bar_objdesc
>
> obj_id 0 obj_name '5525' obj_type 'L' act_id 0 act_type 2 act_status 0
>
> act_start '2015-11-11 15:53:58' act_end '2015-11-10 21:39:49'
>
> ins_time 0 rsam_time 0 seal_time 1447187988 prev_seal_time 1447155921
> level
> 0 copyid hi:lo 2866:2015011010 req_act_id 7150
>
> logstream 0 est_pages hi:lo 0:0 first_log 0 chpt_log 5525 last_log 0
>
> partial_flag 0 do_query 0 ins_sm_id 0 ins_sm_name ''
>
> ins_verify 0 ins_verify_date '' restore order 0:0
>
> objInfo ''
>
> retry 0 in_catalog 1 in_bootfile 0 child_pid 0 child_state 0
>
> bkup_host '' backup_order 0
> * 2015-11-11 15:53:58 180 180 (-43258) ON-Bar could not get logical log
> 5525 fro*
> *m the storage manager.*
> 2015-11-11 15:54:21 180 180 (-43168) ON-Bar suspended the logical restore
> on l
> og 5525 (expected to
> restore to 2147483647).
> 2015-11-11 15:54:27 180 180 C:\\\\PROGRA~1\\\\IBM\\\\Informix\\\\11.70\\\\bin\\\\onbar_d
> complet
> e, returning 100 (0x64)
>
> I tried manually to PITR to log 5525 using either of the specifications(set
> OB2BARLIST), but again onbar can't find the logs :(
>
> So my question is: besides my attempts to reschedule the full backup in a
> time with lowest workload possible, is there something else I can try to do
> to prevent this situation happens in the future?
>
> Or this is limitation of the Data protector integration for Informix?
>
> Thanks in advance for any comments!
>
> --001a114daacc33c074052445ca1e
>
>
>
*******************************************************************************
> Forum Note: Use "Reply" to post a response in the discussion forum.
>
I forgot to mention that we take whole-system backups, sorry.
So as they are based on a checkpoint time including the logs,maybe these
two logs were needed to satisfy the checkpoint.
I am aware that whole system backups do not require to restore additional
logs,but it is written in the docs as option for roll forward recovery
if/when needed.
So basically I can't remove the Logs being backed up by the whole backup
specification,because it will fail to complete.
So basically I can't do much if I understand it right?
Thanks for your time!
BR,
Lyubomir
On Nov 11, 2015 6:44 PM, "Alexandre Marini" <alexandre@briug.org> wrote:
> Hello.
> This kind of errors are common to DP, if you are using 2 separate jobs (one
> for archives, and the other for logs through alarmprogram).
> I just don't understand what job did backup the 2 logs you've mentioned.
> Did
> any onbar process hang in your filesystem?? If so, were they killed
> manually???
> As the error log description says, the next logical log backup should
> automatically include the 2 (or more) missed logs. No magic.
>
> Just make sure your archive job in DP does not have a L checked in your
> Resource type configuration, and also the opposite (your logical logs job
> must
> only have the L checkbox marked, in its Resource Type configuration).
>
> That should separate and not start logical logs backups by the archive job.
>
> Alexandre Marini
> IBM Informix Certified Professional v10 / v11.50 / v11.70 / v12.10
>
> IBM Information Management Informix Technical Professional
>
> IBM Certified Developer - Informix Genero
> BRIUG website administrator
> Informix independent consultant
>
> > To: ids@iiug.org
> > From: lgrigorlu1@gmail.com
> > Subject: onbar Logical restore cannot complete because .... [36046]
> > Date: Wed, 11 Nov 2015 10:50:51 -0500
> >
> > Hello all,
> >
> > OS: Windows Server 2008 R2
> > DB: Informix IDS 11.70 FC8GE
> > SM: HP Data Protector 9
> >
> > We have two separate backup specifications - one for full backups, one
> for
> > log backups. The instance is configured with environment variable
> parameter
> > OB2BARLIST poining to the log specification in the instance .cmd console
> > file and in ALARMPROGRAM.
> > The full backups are being run by Scheduler of Data protector each day
> > using the Full backup specifications. The log backups are being triggered
> > by ALARMPROGRAM when a log fills up.
> >
> > During our monthly restore tests, we've hit an error which prevents us
> from
> > performing Logical restore of the instance. The problem is that onbar
> > cannot find two logs from the Storage manager, which are the first two
> logs
> > after the time Full backup was taken. The problem is that when Full
> backup
> > started, ALARMPROGRAM started trying to backup current log log 5525 (the
> > current log by the time when Full started). Then, another log 5526 was
> > filled up and also was backed up by the Full backup session. In the
> > meantime, ALARMPROGRAM caused error 43395 An logical log backup is
> already
> > running - cannot start another.
> >
> > So at the end we have the Full backup specification holding all the
> > dbspaces, critical files, and logs 5525 and 5526 and the Log
> specification
> > holding up all logical logs besides these.
> >
> > The Physical restore (onbar -r -p) using the Full backup specification
> > succeeds without worries, but after when we try logical restore (onbar -r
> > -l) onbar complains it can't find log 5525:
> >
> > 2015-11-11 15:19:02 180 180 Begin reserved pages restore (level 0 of
> > rootdbs,
> > Storage Manager copy ID: 2866 2015011010).
> > 2015-11-11 15:19:03 180 180 Completed reserved pages restore (level 0 of
> > rootd
> > bs).
> > 2015-11-11 15:19:04 180 180 Successfully connected to Storage Manager.
> > 2015-11-11 15:19:39 180 180 Begin cold level 0 restore rootdbs (Storage
> > Manage
> > r copy ID: 2866 2015011010).
> > 2015-11-11 15:19:59 180 180 Completed cold level 0 restore rootdbs.
> > 2015-11-11 15:20:35 180 180 Begin cold level 0 restore
> > ona2aqifx030p0dbs01 (St
> > orage Manager copy ID: 2866 2015011010).
> > 2015-11-11 15:52:02 180 180 Completed cold level 0 restore
> > ona2aqifx030p0dbs01
> > ..
> > 2015-11-11 15:52:38 180 180 Begin cold level 0 restore sbspace01 (Storage
> > Mana
> > ger copy ID: 2866 2015011010).
> > 2015-11-11 15:52:40 180 180 Completed cold level 0 restore sbspace01.
> > 2015-11-11 15:53:16 180 180 Begin cold level 0 restore logdbs01 (Storage
> > Manag
> > er copy ID: 2866 2015011010).
> > 2015-11-11 15:53:18 180 180 Completed cold level 0 restore logdbs01.
> > 2015-11-11 15:53:53 180 180 Begin cold level 0 restore physdbs (Storage
> > Manage
> > r copy ID: 2866 2015011010).
> > 2015-11-11 15:53:56 180 180 Completed cold level 0 restore physdbs.
> > 2015-11-11 15:53:58 180 180 Successfully connected to Storage Manager.
> > 2015-11-11 15:53:58 180 180 XBSA Error: (BSAGetObject) Backup object does
> > not
> > exist in Storage Manager.
> > 2015-11-11 15:53:58 180 180 Object information: bar_objdesc
> >
> > obj_id 0 obj_name '5525' obj_type 'L' act_id 0 act_type 2 act_status 0
> >
> > act_start '2015-11-11 15:53:58' act_end '2015-11-10 21:39:49'
> >
> > ins_time 0 rsam_time 0 seal_time 1447187988 prev_seal_time 1447155921
> > level
> > 0 copyid hi:lo 2866:2015011010 req_act_id 7150
> >
> > logstream 0 est_pages hi:lo 0:0 first_log 0 chpt_log 5525 last_log 0
> >
> > partial_flag 0 do_query 0 ins_sm_id 0 ins_sm_name ''
> >
> > ins_verify 0 ins_verify_date '' restore order 0:0
> >
> > objInfo ''
> >
> > retry 0 in_catalog 1 in_bootfile 0 child_pid 0 child_state 0
> >
> > bkup_host '' backup_order 0
> > * 2015-11-11 15:53:58 180 180 (-43258) ON-Bar could not get logical log
> > 5525 fro*
> > *m the storage manager.*
> > 2015-11-11 15:54:21 180 180 (-43168) ON-Bar suspended the logical restore
> > on l
> > og 5525 (expected to
> > restore to 2147483647).
> > 2015-11-11 15:54:27 180 180 C:\\\\PROGRA~1\\\\IBM\\\\Informix\\\\11.70\\\\bin\\\\onbar_d
> > complet
> > e, returning 100 (0x64)
> >
> > I tried manually to PITR to log 5525 using either of the
> specifications(set
> > OB2BARLIST), but again onbar can't find the logs :(
> >
> > So my question is: besides my attempts to reschedule the full backup in a
> > time with lowest workload possible, is there something else I can try to
> do
> > to prevent this situation happens in the future?
> >
> > Or this is limitation of the Data protector integration for Informix?
> >
> > Thanks in advance for any comments!
> >
> > --001a114daacc33c074052445ca1e
> >
> >
> >
>
>
*******************************************************************************
> > Forum Note: Use "Reply" to post a response in the discussion forum.
> >
>
>
>
>
*******************************************************************************
> Forum Note: Use "Reply" to post a response in the discussion forum.
>
>
--001a114e4d7cd96a2d0524487e7e
One question:
If I change the backups from Whole system (with -w flag) to Parralel
(without -w flag) backup, and I differentiate the specifications so that
logs only go to log specification, and not the full backup specification,
will this work correctly?
I mean, during L0 backup, will Informix require to backup the current log
together with the L0 backup, or this can be done afterwards with Log
backups from ALARMPROGRAM? Because as far as I know we can't have two
backups running at a time :)
Thanks a lot for the information!
2015-11-11 21:04 GMT+02:00 Lyubomir Grigorov <lgrigorlu1@gmail.com>:
> I forgot to mention that we take whole-system backups, sorry.
>
> So as they are based on a checkpoint time including the logs,maybe these
> two logs were needed to satisfy the checkpoint.
>
> I am aware that whole system backups do not require to restore additional
> logs,but it is written in the docs as option for roll forward recovery
> if/when needed.
>
> So basically I can't remove the Logs being backed up by the whole backup
> specification,because it will fail to complete.
>
> So basically I can't do much if I understand it right?
>
> Thanks for your time!
>
> BR,
> Lyubomir
> On Nov 11, 2015 6:44 PM, "Alexandre Marini" <alexandre@briug.org> wrote:
>
> > Hello.
> > This kind of errors are common to DP, if you are using 2 separate jobs
> (one
> > for archives, and the other for logs through alarmprogram).
> > I just don't understand what job did backup the 2 logs you've mentioned.
> > Did
> > any onbar process hang in your filesystem?? If so, were they killed
> > manually???
> > As the error log description says, the next logical log backup should
> > automatically include the 2 (or more) missed logs. No magic.
> >
> > Just make sure your archive job in DP does not have a L checked in your
> > Resource type configuration, and also the opposite (your logical logs job
> > must
> > only have the L checkbox marked, in its Resource Type configuration).
> >
> > That should separate and not start logical logs backups by the archive
> job.
> >
> > Alexandre Marini
> > IBM Informix Certified Professional v10 / v11.50 / v11.70 / v12.10
> >
> > IBM Information Management Informix Technical Professional
> >
> > IBM Certified Developer - Informix Genero
> > BRIUG website administrator
> > Informix independent consultant
> >
> > > To: ids@iiug.org
> > > From: lgrigorlu1@gmail.com
> > > Subject: onbar Logical restore cannot complete because .... [36046]
> > > Date: Wed, 11 Nov 2015 10:50:51 -0500
> > >
> > > Hello all,
> > >
> > > OS: Windows Server 2008 R2
> > > DB: Informix IDS 11.70 FC8GE
> > > SM: HP Data Protector 9
> > >
> > > We have two separate backup specifications - one for full backups, one
> > for
> > > log backups. The instance is configured with environment variable
> > parameter
> > > OB2BARLIST poining to the log specification in the instance .cmd
> console
> > > file and in ALARMPROGRAM.
> > > The full backups are being run by Scheduler of Data protector each day
> > > using the Full backup specifications. The log backups are being
> triggered
> > > by ALARMPROGRAM when a log fills up.
> > >
> > > During our monthly restore tests, we've hit an error which prevents us
> > from
> > > performing Logical restore of the instance. The problem is that onbar
> > > cannot find two logs from the Storage manager, which are the first two
> > logs
> > > after the time Full backup was taken. The problem is that when Full
> > backup
> > > started, ALARMPROGRAM started trying to backup current log log 5525
> (the
> > > current log by the time when Full started). Then, another log 5526 was
> > > filled up and also was backed up by the Full backup session. In the
> > > meantime, ALARMPROGRAM caused error 43395 An logical log backup is
> > already
> > > running - cannot start another.
> > >
> > > So at the end we have the Full backup specification holding all the
> > > dbspaces, critical files, and logs 5525 and 5526 and the Log
> > specification
> > > holding up all logical logs besides these.
> > >
> > > The Physical restore (onbar -r -p) using the Full backup specification
> > > succeeds without worries, but after when we try logical restore (onbar
> -r
> > > -l) onbar complains it can't find log 5525:
> > >
> > > 2015-11-11 15:19:02 180 180 Begin reserved pages restore (level 0 of
> > > rootdbs,
> > > Storage Manager copy ID: 2866 2015011010).
> > > 2015-11-11 15:19:03 180 180 Completed reserved pages restore (level 0
> of
> > > rootd
> > > bs).
> > > 2015-11-11 15:19:04 180 180 Successfully connected to Storage Manager.
> > > 2015-11-11 15:19:39 180 180 Begin cold level 0 restore rootdbs (Storage
> > > Manage
> > > r copy ID: 2866 2015011010).
> > > 2015-11-11 15:19:59 180 180 Completed cold level 0 restore rootdbs.
> > > 2015-11-11 15:20:35 180 180 Begin cold level 0 restore
> > > ona2aqifx030p0dbs01 (St
> > > orage Manager copy ID: 2866 2015011010).
> > > 2015-11-11 15:52:02 180 180 Completed cold level 0 restore
> > > ona2aqifx030p0dbs01
> > > ..
> > > 2015-11-11 15:52:38 180 180 Begin cold level 0 restore sbspace01
> (Storage
> > > Mana
> > > ger copy ID: 2866 2015011010).
> > > 2015-11-11 15:52:40 180 180 Completed cold level 0 restore sbspace01.
> > > 2015-11-11 15:53:16 180 180 Begin cold level 0 restore logdbs01
> (Storage
> > > Manag
> > > er copy ID: 2866 2015011010).
> > > 2015-11-11 15:53:18 180 180 Completed cold level 0 restore logdbs01.
> > > 2015-11-11 15:53:53 180 180 Begin cold level 0 restore physdbs (Storage
> > > Manage
> > > r copy ID: 2866 2015011010).
> > > 2015-11-11 15:53:56 180 180 Completed cold level 0 restore physdbs.
> > > 2015-11-11 15:53:58 180 180 Successfully connected to Storage Manager.
> > > 2015-11-11 15:53:58 180 180 XBSA Error: (BSAGetObject) Backup object
> does
> > > not
> > > exist in Storage Manager.
> > > 2015-11-11 15:53:58 180 180 Object information: bar_objdesc
> > >
> > > obj_id 0 obj_name '5525' obj_type 'L' act_id 0 act_type 2 act_status 0
> > >
> > > act_start '2015-11-11 15:53:58' act_end '2015-11-10 21:39:49'
> > >
> > > ins_time 0 rsam_time 0 seal_time 1447187988 prev_seal_time 1447155921
> > > level
> > > 0 copyid hi:lo 2866:2015011010 req_act_id 7150
> > >
> > > logstream 0 est_pages hi:lo 0:0 first_log 0 chpt_log 5525 last_log 0
> > >
> > > partial_flag 0 do_query 0 ins_sm_id 0 ins_sm_name ''
> > >
> > > ins_verify 0 ins_verify_date '' restore order 0:0
> > >
> > > objInfo ''
> > >
> > > retry 0 in_catalog 1 in_bootfile 0 child_pid 0 child_state 0
> > >
> > > bkup_host '' backup_order 0
> > > * 2015-11-11 15:53:58 180 180 (-43258) ON-Bar could not get logical log
> > > 5525 fro*
> > > *m the storage manager.*
> > > 2015-11-11 15:54:21 180 180 (-43168) ON-Bar suspended the logical
> restore
> > > on l
> > > og 5525 (expected to
> > > restore to 2147483647).
> > > 2015-11-11 15:54:27 180 180 C:\\\\PROGRA~1\\\\IBM\\\\Informix\\\\11.70\\\\bin\\\\onbar_d
> > > complet
> > > e, returning 100 (0x64)
> > >
> > > I trie
Hello, Lyubomir.
That's the idea to test ,and make sure it works as expected.
I just wanted to ensure your whole system restore worked, finding those 2
missing logs first.
If you want to discard that test, and change the configs, ok.
Best regards.
Alexandre Marini
IBM Informix Certified Professional v10 / v11.50 / v11.70 / v12.10
IBM Information Management Informix Technical Professional
IBM Certified Developer - Informix Genero
BRIUG website administrator
Informix independent consultant
> To: ids@iiug.org
> From: lgrigorlu1@gmail.com
> Subject: Re: onbar Logical restore cannot complete beca.... [36050]
> Date: Thu, 12 Nov 2015 02:38:05 -0500
>
> One question:
>
> If I change the backups from Whole system (with -w flag) to Parralel
> (without -w flag) backup, and I differentiate the specifications so that
> logs only go to log specification, and not the full backup specification,
> will this work correctly?
>
> I mean, during L0 backup, will Informix require to backup the current log
> together with the L0 backup, or this can be done afterwards with Log
> backups from ALARMPROGRAM? Because as far as I know we can't have two
> backups running at a time :)
>
> Thanks a lot for the information!
>
> 2015-11-11 21:04 GMT+02:00 Lyubomir Grigorov <lgrigorlu1@gmail.com>:
>
> > I forgot to mention that we take whole-system backups, sorry.
> >
> > So as they are based on a checkpoint time including the logs,maybe these
> > two logs were needed to satisfy the checkpoint.
> >
> > I am aware that whole system backups do not require to restore additional
> > logs,but it is written in the docs as option for roll forward recovery
> > if/when needed.
> >
> > So basically I can't remove the Logs being backed up by the whole backup
> > specification,because it will fail to complete.
> >
> > So basically I can't do much if I understand it right?
> >
> > Thanks for your time!
> >
> > BR,
> > Lyubomir
> > On Nov 11, 2015 6:44 PM, "Alexandre Marini" <alexandre@briug.org> wrote:
> >
> > > Hello.
> > > This kind of errors are common to DP, if you are using 2 separate jobs
> > (one
> > > for archives, and the other for logs through alarmprogram).
> > > I just don't understand what job did backup the 2 logs you've mentioned.
> > > Did
> > > any onbar process hang in your filesystem?? If so, were they killed
> > > manually???
> > > As the error log description says, the next logical log backup should
> > > automatically include the 2 (or more) missed logs. No magic.
> > >
> > > Just make sure your archive job in DP does not have a L checked in your
> > > Resource type configuration, and also the opposite (your logical logs job
> > > must
> > > only have the L checkbox marked, in its Resource Type configuration).
> > >
> > > That should separate and not start logical logs backups by the archive
> > job.
> > >
> > > Alexandre Marini
> > > IBM Informix Certified Professional v10 / v11.50 / v11.70 / v12.10
> > >
> > > IBM Information Management Informix Technical Professional
> > >
> > > IBM Certified Developer - Informix Genero
> > > BRIUG website administrator
> > > Informix independent consultant
> > >
> > > > To: ids@iiug.org
> > > > From: lgrigorlu1@gmail.com
> > > > Subject: onbar Logical restore cannot complete because .... [36046]
> > > > Date: Wed, 11 Nov 2015 10:50:51 -0500
> > > >
> > > > Hello all,
> > > >
> > > > OS: Windows Server 2008 R2
> > > > DB: Informix IDS 11.70 FC8GE
> > > > SM: HP Data Protector 9
> > > >
> > > > We have two separate backup specifications - one for full backups, one
> > > for
> > > > log backups. The instance is configured with environment variable
> > > parameter
> > > > OB2BARLIST poining to the log specification in the instance .cmd
> > console
> > > > file and in ALARMPROGRAM.
> > > > The full backups are being run by Scheduler of Data protector each day
> > > > using the Full backup specifications. The log backups are being
> > triggered
> > > > by ALARMPROGRAM when a log fills up.
> > > >
> > > > During our monthly restore tests, we've hit an error which prevents us
> > > from
> > > > performing Logical restore of the instance. The problem is that onbar
> > > > cannot find two logs from the Storage manager, which are the first two
> > > logs
> > > > after the time Full backup was taken. The problem is that when Full
> > > backup
> > > > started, ALARMPROGRAM started trying to backup current log log 5525
> > (the
> > > > current log by the time when Full started). Then, another log 5526 was
> > > > filled up and also was backed up by the Full backup session. In the
> > > > meantime, ALARMPROGRAM caused error 43395 An logical log backup is
> > > already
> > > > running - cannot start another.
> > > >
> > > > So at the end we have the Full backup specification holding all the
> > > > dbspaces, critical files, and logs 5525 and 5526 and the Log
> > > specification
> > > > holding up all logical logs besides these.
> > > >
> > > > The Physical restore (onbar -r -p) using the Full backup specification
> > > > succeeds without worries, but after when we try logical restore (onbar
> > -r
> > > > -l) onbar complains it can't find log 5525:
> > > >
> > > > 2015-11-11 15:19:02 180 180 Begin reserved pages restore (level 0 of
> > > > rootdbs,
> > > > Storage Manager copy ID: 2866 2015011010).
> > > > 2015-11-11 15:19:03 180 180 Completed reserved pages restore (level 0
> > of
> > > > rootd
> > > > bs).
> > > > 2015-11-11 15:19:04 180 180 Successfully connected to Storage Manager.
> > > > 2015-11-11 15:19:39 180 180 Begin cold level 0 restore rootdbs (Storage
> > > > Manage
> > > > r copy ID: 2866 2015011010).
> > > > 2015-11-11 15:19:59 180 180 Completed cold level 0 restore rootdbs.
> > > > 2015-11-11 15:20:35 180 180 Begin cold level 0 restore
> > > > ona2aqifx030p0dbs01 (St
> > > > orage Manager copy ID: 2866 2015011010).
> > > > 2015-11-11 15:52:02 180 180 Completed cold level 0 restore
> > > > ona2aqifx030p0dbs01
> > > > ..
> > > > 2015-11-11 15:52:38 180 180 Begin cold level 0 restore sbspace01
> > (Storage
> > > > Mana
> > > > ger copy ID: 2866 2015011010).
> > > > 2015-11-11 15:52:40 180 180 Completed cold level 0 restore sbspace01.
> > > > 2015-11-11 15:53:16 180 180 Begin cold level 0 restore logdbs01
> > (Storage
> > > > Manag
> > > > er copy ID: 2866 2015011010).
> > > > 2015-11-11 15:53:18 180 180 Completed cold level 0 restore logdbs01.
> > > > 2015-11-11 15:53:53 180 180 Begin cold level 0 restore physdbs (Storage
> > > > Manage
> > > > r copy ID: 2866 2015011010).
> > > > 2015-11-11 15:53:56 180 180 Completed cold level 0 restore physdbs.
> > > > 2015-11-11 15:53:58 180 180 Successfully connected to Storage Manager.
> > > > 2015-11-11 15:53:58 180 180 XBSA Error: (BSAGetObject) Backup object
> > does
> > > > not
> > > > exist in Storage Manager.
> > > > 2015-11-11 15:53:58 180 180 Object information: bar_objdesc
> > > >
> > > > obj_id 0 obj_name '5525' obj_type 'L' act_id 0 act_type 2 act_status 0
> > > >
> > > > act_start '2015-11-11 15:53:58' act_end '
Yes I will reconfigure one instance and will test it.
As for your question: yes,the physical recovery of the whole system backup
restores fine, in message and onbar logs you see that cold restore of
dbspaces complete and none of the logs included in the backup are
mentioned. And then when you try any form of roll-forward recovery fails
complaining that needed log/logs can't be found on SM.
Regards,
Lyubomir Grigorov
On Nov 12, 2015 12:50 PM, "Alexandre Marini" <alexandre@briug.org> wrote:
> Hello, Lyubomir.
> That's the idea to test ,and make sure it works as expected.
>
> I just wanted to ensure your whole system restore worked, finding those 2
> missing logs first.
> If you want to discard that test, and change the configs, ok.
>
> Best regards.
>
> Alexandre Marini
> IBM Informix Certified Professional v10 / v11.50 / v11.70 / v12.10
>
> IBM Information Management Informix Technical Professional
>
> IBM Certified Developer - Informix Genero
> BRIUG website administrator
> Informix independent consultant
>
> > To: ids@iiug.org
> > From: lgrigorlu1@gmail.com
> > Subject: Re: onbar Logical restore cannot complete beca.... [36050]
> > Date: Thu, 12 Nov 2015 02:38:05 -0500
> >
> > One question:
> >
> > If I change the backups from Whole system (with -w flag) to Parralel
> > (without -w flag) backup, and I differentiate the specifications so that
> > logs only go to log specification, and not the full backup specification,
> > will this work correctly?
> >
> > I mean, during L0 backup, will Informix require to backup the current log
> > together with the L0 backup, or this can be done afterwards with Log
> > backups from ALARMPROGRAM? Because as far as I know we can't have two
> > backups running at a time :)
> >
> > Thanks a lot for the information!
> >
> > 2015-11-11 21:04 GMT+02:00 Lyubomir Grigorov <lgrigorlu1@gmail.com>:
> >
> > > I forgot to mention that we take whole-system backups, sorry.
> > >
> > > So as they are based on a checkpoint time including the logs,maybe
> these
> > > two logs were needed to satisfy the checkpoint.
> > >
> > > I am aware that whole system backups do not require to restore
> additional
> > > logs,but it is written in the docs as option for roll forward recovery
> > > if/when needed.
> > >
> > > So basically I can't remove the Logs being backed up by the whole
> backup
> > > specification,because it will fail to complete.
> > >
> > > So basically I can't do much if I understand it right?
> > >
> > > Thanks for your time!
> > >
> > > BR,
> > > Lyubomir
> > > On Nov 11, 2015 6:44 PM, "Alexandre Marini" <alexandre@briug.org>
> wrote:
> > >
> > > > Hello.
> > > > This kind of errors are common to DP, if you are using 2 separate
> jobs
> > > (one
> > > > for archives, and the other for logs through alarmprogram).
> > > > I just don't understand what job did backup the 2 logs you've
> mentioned.
> > > > Did
> > > > any onbar process hang in your filesystem?? If so, were they killed
> > > > manually???
> > > > As the error log description says, the next logical log backup should
> > > > automatically include the 2 (or more) missed logs. No magic.
> > > >
> > > > Just make sure your archive job in DP does not have a L checked in
> your
> > > > Resource type configuration, and also the opposite (your logical logs
> job
> > > > must
> > > > only have the L checkbox marked, in its Resource Type configuration).
> > > >
> > > > That should separate and not start logical logs backups by the
> archive
> > > job.
> > > >
> > > > Alexandre Marini
> > > > IBM Informix Certified Professional v10 / v11.50 / v11.70 / v12.10
> > > >
> > > > IBM Information Management Informix Technical Professional
> > > >
> > > > IBM Certified Developer - Informix Genero
> > > > BRIUG website administrator
> > > > Informix independent consultant
> > > >
> > > > > To: ids@iiug.org
> > > > > From: lgrigorlu1@gmail.com
> > > > > Subject: onbar Logical restore cannot complete because .... [36046]
> > > > > Date: Wed, 11 Nov 2015 10:50:51 -0500
> > > > >
> > > > > Hello all,
> > > > >
> > > > > OS: Windows Server 2008 R2
> > > > > DB: Informix IDS 11.70 FC8GE
> > > > > SM: HP Data Protector 9
> > > > >
> > > > > We have two separate backup specifications - one for full backups,
> one
> > > > for
> > > > > log backups. The instance is configured with environment variable
> > > > parameter
> > > > > OB2BARLIST poining to the log specification in the instance .cmd
> > > console
> > > > > file and in ALARMPROGRAM.
> > > > > The full backups are being run by Scheduler of Data protector each
> day
> > > > > using the Full backup specifications. The log backups are being
> > > triggered
> > > > > by ALARMPROGRAM when a log fills up.
> > > > >
> > > > > During our monthly restore tests, we've hit an error which
> prevents us
> > > > from
> > > > > performing Logical restore of the instance. The problem is that
> onbar> > > > > cannot find two logs from the Storage manager, which are the first
> two
> > > > logs
> > > > > after the time Full backup was taken. The problem is that when Full
> > > > backup
> > > > > started, ALARMPROGRAM started trying to backup current log log 5525
> > > (the
> > > > > current log by the time when Full started). Then, another log 5526
> was
> > > > > filled up and also was backed up by the Full backup session. In the
> > > > > meantime, ALARMPROGRAM caused error 43395 An logical log backup is
> > > > already
> > > > > running - cannot start another.
> > > > >
> > > > > So at the end we have the Full backup specification holding all the
> > > > > dbspaces, critical files, and logs 5525 and 5526 and the Log
> > > > specification
> > > > > holding up all logical logs besides these.
> > > > >
> > > > > The Physical restore (onbar -r -p) using the Full backup
> specification
> > > > > succeeds without worries, but after when we try logical restore
> (onbar
> > > -r
> > > > > -l) onbar complains it can't find log 5525:
> > > > >
> > > > > 2015-11-11 15:19:02 180 180 Begin reserved pages restore (level 0
> of
> > > > > rootdbs,
> > > > > Storage Manager copy ID: 2866 2015011010).
> > > > > 2015-11-11 15:19:03 180 180 Completed reserved pages restore
> (level 0
> > > of
> > > > > rootd
> > > > > bs).
> > > > > 2015-11-11 15:19:04 180 180 Successfully connected to Storage
> Manager.
> > > > > 2015-11-11 15:19:39 180 180 Begin cold level 0 restore rootdbs
> (Storage
> > > > > Manage
> > > > > r copy ID: 2866 2015011010).
> > > > > 2015-11-11 15:19:59 180 180 Completed cold level 0 restore rootdbs.
> > > > > 2015-11-11 15:20:35 180 180 Begin cold level 0 restore
> > > > > ona2aqifx030p0dbs01 (St
> > > > > orage Manager copy ID: 2866 2015011010).
> > > > > 2015-11-11 15:52:02 180 180 Completed cold level 0 restore
> > > > > ona2aqifx030p0dbs01
> > > > > ..
> > > > > 2015-11-11 15:52:38 180 180 Begin cold level 0 restore sbspace01
> > > (Storage
> > > > > Mana
> > > > > ger copy ID: 2866 2015011010).@@
Ok I see.
So change your full backup (archive) to non whole system, and everything must
work fine, even if any logical-log backup try to run and fails, the next one
will proceed anyway.
Best Regards.
Alexandre Marini
IBM Informix Certified Professional v10 / v11.50 / v11.70 / v12.10
IBM Information Management Informix Technical Professional
IBM Certified Developer - Informix Genero
BRIUG website administrator
Informix independent consultant
> To: ids@iiug.org
> From: lgrigorlu1@gmail.com
> Subject: RE: onbar Logical restore cannot complete beca.... [36053]
> Date: Thu, 12 Nov 2015 07:40:26 -0500
>
> Yes I will reconfigure one instance and will test it.
>
> As for your question: yes,the physical recovery of the whole system backup
> restores fine, in message and onbar logs you see that cold restore of
> dbspaces complete and none of the logs included in the backup are
> mentioned. And then when you try any form of roll-forward recovery fails
> complaining that needed log/logs can't be found on SM.
>
> Regards,
> Lyubomir Grigorov
> On Nov 12, 2015 12:50 PM, "Alexandre Marini" <alexandre@briug.org> wrote:
>
> > Hello, Lyubomir.
> > That's the idea to test ,and make sure it works as expected.
> >
> > I just wanted to ensure your whole system restore worked, finding those 2
> > missing logs first.
> > If you want to discard that test, and change the configs, ok.
> >
> > Best regards.
> >
> > Alexandre Marini
> > IBM Informix Certified Professional v10 / v11.50 / v11.70 / v12.10
> >
> > IBM Information Management Informix Technical Professional
> >
> > IBM Certified Developer - Informix Genero
> > BRIUG website administrator
> > Informix independent consultant
> >
> > > To: ids@iiug.org
> > > From: lgrigorlu1@gmail.com
> > > Subject: Re: onbar Logical restore cannot complete beca.... [36050]
> > > Date: Thu, 12 Nov 2015 02:38:05 -0500
> > >
> > > One question:
> > >
> > > If I change the backups from Whole system (with -w flag) to Parralel
> > > (without -w flag) backup, and I differentiate the specifications so that
> > > logs only go to log specification, and not the full backup specification,
> > > will this work correctly?
> > >
> > > I mean, during L0 backup, will Informix require to backup the current log
> > > together with the L0 backup, or this can be done afterwards with Log
> > > backups from ALARMPROGRAM? Because as far as I know we can't have two
> > > backups running at a time :)
> > >
> > > Thanks a lot for the information!
> > >
> > > 2015-11-11 21:04 GMT+02:00 Lyubomir Grigorov <lgrigorlu1@gmail.com>:
> > >
> > > > I forgot to mention that we take whole-system backups, sorry.
> > > >
> > > > So as they are based on a checkpoint time including the logs,maybe
> > these
> > > > two logs were needed to satisfy the checkpoint.
> > > >
> > > > I am aware that whole system backups do not require to restore
> > additional
> > > > logs,but it is written in the docs as option for roll forward recovery
> > > > if/when needed.
> > > >
> > > > So basically I can't remove the Logs being backed up by the whole
> > backup
> > > > specification,because it will fail to complete.
> > > >
> > > > So basically I can't do much if I understand it right?
> > > >
> > > > Thanks for your time!
> > > >
> > > > BR,
> > > > Lyubomir
> > > > On Nov 11, 2015 6:44 PM, "Alexandre Marini" <alexandre@briug.org>
> > wrote:
> > > >
> > > > > Hello.
> > > > > This kind of errors are common to DP, if you are using 2 separate
> > jobs
> > > > (one
> > > > > for archives, and the other for logs through alarmprogram).
> > > > > I just don't understand what job did backup the 2 logs you've
> > mentioned.
> > > > > Did
> > > > > any onbar process hang in your filesystem?? If so, were they killed
> > > > > manually???
> > > > > As the error log description says, the next logical log backup should
> > > > > automatically include the 2 (or more) missed logs. No magic.
> > > > >
> > > > > Just make sure your archive job in DP does not have a L checked in
> > your
> > > > > Resource type configuration, and also the opposite (your logical logs
> > job
> > > > > must
> > > > > only have the L checkbox marked, in its Resource Type configuration).
> > > > >
> > > > > That should separate and not start logical logs backups by the
> > archive
> > > > job.
> > > > >
> > > > > Alexandre Marini
> > > > > IBM Informix Certified Professional v10 / v11.50 / v11.70 / v12.10
> > > > >
> > > > > IBM Information Management Informix Technical Professional
> > > > >
> > > > > IBM Certified Developer - Informix Genero
> > > > > BRIUG website administrator
> > > > > Informix independent consultant
> > > > >
> > > > > > To: ids@iiug.org
> > > > > > From: lgrigorlu1@gmail.com
> > > > > > Subject: onbar Logical restore cannot complete because .... [36046]
> > > > > > Date: Wed, 11 Nov 2015 10:50:51 -0500
> > > > > >
> > > > > > Hello all,
> > > > > >
> > > > > > OS: Windows Server 2008 R2
> > > > > > DB: Informix IDS 11.70 FC8GE
> > > > > > SM: HP Data Protector 9
> > > > > >
> > > > > > We have two separate backup specifications - one for full backups,
> > one
> > > > > for
> > > > > > log backups. The instance is configured with environment variable
> > > > > parameter
> > > > > > OB2BARLIST poining to the log specification in the instance .cmd
> > > > console
> > > > > > file and in ALARMPROGRAM.
> > > > > > The full backups are being run by Scheduler of Data protector each
> > day
> > > > > > using the Full backup specifications. The log backups are being
> > > > triggered
> > > > > > by ALARMPROGRAM when a log fills up.
> > > > > >
> > > > > > During our monthly restore tests, we've hit an error which
> > prevents us
> > > > > from
> > > > > > performing Logical restore of the instance. The problem is that
> > onbar> > > > > > cannot find two logs from the Storage manager, which are the first
> > two
> > > > > logs
> > > > > > after the time Full backup was taken. The problem is that when Full
> > > > > backup
> > > > > > started, ALARMPROGRAM started trying to backup current log log 5525
> > > > (the
> > > > > > current log by the time when Full started). Then, another log 5526
> > was
> > > > > > filled up and also was backed up by the Full backup session. In the
> > > > > > meantime, ALARMPROGRAM caused error 43395 An logical log backup is
> > > > > already
> > > > > > running - cannot start another.
> > > > > >
> > > > > > So at the end we have the Full backup specification holding all the
> > > > > > dbspaces, critical files, and logs 5525 and 5526 and the Log
> > > > > specification
> > > > > > holding up all logical logs besides these.
> > > > > >
> > > > > > The Physical restore (onbar -r -p) using the Full backup
> > specification
> > > > > > succeeds without worries, but after when we try logical restore
> > (onbar
> > > > -r
> > > > > > -l) onbar complains it can't find log 5525:
> > > > > >
> > > > > > 2015-11-11 15:19:02 180 180 Begin reserved pages restore (
Some years ago when we use to still back up to tape, in the batch file that started the full backup, we would first rename '%informixdir%\\\\etc\\\\log_full.bat' to something else so that no triggered log backups could happen during the full backup (I think it gets called by the alarmprogram.bat). Once the full backup completed we'd just rename it back to log_full.bat and the next full log would get all the logs again. Would that work in your situation? Regards, Bryce Stenberg.
Hi, I've finally found out what's the problem: the difference in the SERVERNUM between source instance and target instance(the one that is restored). On source instance servernum is 0, and on target machine, if you have another instance with serverrnum 0 like I did, you need to change servernum to other value. That's fine, but in the case of logical restore, obviously the Logs are being "marked" with the servernum in Storage manager IDB, thus when instance with different servernum requests a log, it fails with error: 5880 5880 (-43258) ON-Bar could not get logical log 5432 from the storage manager. 5880 5880 (-43168) ON-Bar suspended the logical restore on log 5432 (expected to restore to 2147483647). It doesn't matter if the backup has been taken with or without Whole-System (-w) flag. Once I change servernum to 0 like on source, the restore (Logical and/or PITR) completes just fine. I've also tried to change in ixbar file 4th column (which is servernum value), but again this didn't help and the needed logs cannot be found on SM. I've also looked up the bar_ tables in sysutils, but can't find any of these tables to have column like "servernum_id" or similar, so my guess is that SM is marking these logs somehow, or maybe there is some limitation in the Data protector integration. Regards, Lyubomir 2015-11-16 23:48 GMT+02:00 BRYCE STENBERG <bryce@hrnz.co.nz>: > Some years ago when we use to still back up to tape, in the batch file that > started the full backup, we would first rename > '%informixdir%\\\\etc\\\\log_full.bat' to something else so that no triggered log > backups could happen during the full backup (I think it gets called by the > alarmprogram.bat). Once the full backup completed we'd just rename it back > to > log_full.bat and the next full log would get all the logs again. > Would that work in your situation? > > Regards, > Bryce Stenberg. > > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > > --001a11430a368e14ab0524b78eb2
Different storage managers have different strategies for overriding the servernum when you restore so you can restore to a server with a different servernum than the source. I don't remember what storage manager you are using, but check out their manuals for details. Art Art S. Kagel, President and Principal Consultant ASK Database Management www.askdbmgt.com Blog: http://informix-myview.blogspot.com/ Disclaimer: Please keep in mind that my own opinions are my own opinions and do not reflect on the IIUG, nor any other organization with which I am associated either explicitly, implicitly, or by inference. Neither do those opinions reflect those of other individuals affiliated with any entity with which I am affiliated nor those of the entities themselves. On Tue, Nov 17, 2015 at 2:34 AM, Lyubomir Grigorov <lgrigorlu1@gmail.com> wrote: > Hi, > > I've finally found out what's the problem: the difference in the SERVERNUM > between source instance and target instance(the one that is restored). > On source instance servernum is 0, and on target machine, if you have > another instance with serverrnum 0 like I did, you need to change servernum > to other value. That's fine, but in the case of logical restore, obviously > the Logs are being "marked" with the servernum in Storage manager IDB, thus > when instance with different servernum requests a log, it fails with error: > > 5880 5880 (-43258) ON-Bar could not get logical log 5432 from the storage > manager. > 5880 5880 (-43168) ON-Bar suspended the logical restore on log 5432 > (expected to restore to 2147483647). > > It doesn't matter if the backup has been taken with or without Whole-System > (-w) flag. > Once I change servernum to 0 like on source, the restore (Logical and/or > PITR) completes just fine. > > I've also tried to change in ixbar file 4th column (which is servernum > value), but again this didn't help and the needed logs cannot be found on > SM. > > I've also looked up the bar_ tables in sysutils, but can't find any of > these tables to have column like "servernum_id" or similar, so my guess is > that SM is marking these logs somehow, or maybe there is some limitation in > the Data protector integration. > > Regards, > Lyubomir > > 2015-11-16 23:48 GMT+02:00 BRYCE STENBERG <bryce@hrnz.co.nz>: > > > Some years ago when we use to still back up to tape, in the batch file > that > > started the full backup, we would first rename > > '%informixdir%\\\\etc\\\\log_full.bat' to something else so that no triggered > log > > backups could happen during the full backup (I think it gets called by > the > > alarmprogram.bat). Once the full backup completed we'd just rename it > back > > to > > log_full.bat and the next full log would get all the logs again. > > Would that work in your situation? > > > > Regards, > > Bryce Stenberg. > > > > > > > > > > ******************************************************************************* > > Forum Note: Use "Reply" to post a response in the discussion forum. > > > > > > --001a11430a368e14ab0524b78eb2 > > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > > --001a113fe84e414b6f0524bb6a8c
Hello, my friend.
A restore through DP always suppose you use the same onconfig file, so
SERVERNUM is supposed to be always the same as in backup process/catalog.
A friend of mine that works as an HP consultant told me that, for a cold
restore, you should always preserve your original configs. So SERVERNUM must
be the same, I don't know if you can manually change it. I've done a cold
restore, just fine.
In the case that you have multiple residency, each pair of DP jobs must be
configured, one for each instance. That way keeps track of your original
configurations, and the catalog will be fine for each instance you might have.
Hope it helps.
Best regards.
Alexandre Marini
IBM Informix Certified Professional v10 / v11.50 / v11.70 / v12.10
IBM Information Management Informix Technical Professional
IBM Certified Developer - Informix Genero
BRIUG website administrator
Informix independent consultant
> To: ids@iiug.org
> From: lgrigorlu1@gmail.com
> Subject: Re: onbar Logical restore cannot complete because [36068]
> Date: Tue, 17 Nov 2015 02:34:41 -0500
>
> Hi,
>
> I've finally found out what's the problem: the difference in the SERVERNUM
> between source instance and target instance(the one that is restored).
> On source instance servernum is 0, and on target machine, if you have
> another instance with serverrnum 0 like I did, you need to change servernum
> to other value. That's fine, but in the case of logical restore, obviously
> the Logs are being "marked" with the servernum in Storage manager IDB, thus
> when instance with different servernum requests a log, it fails with error:
>
> 5880 5880 (-43258) ON-Bar could not get logical log 5432 from the storage
> manager.
> 5880 5880 (-43168) ON-Bar suspended the logical restore on log 5432
> (expected to restore to 2147483647).
>
> It doesn't matter if the backup has been taken with or without Whole-System
> (-w) flag.
> Once I change servernum to 0 like on source, the restore (Logical and/or
> PITR) completes just fine.
>
> I've also tried to change in ixbar file 4th column (which is servernum
> value), but again this didn't help and the needed logs cannot be found on
> SM.
>
> I've also looked up the bar_ tables in sysutils, but can't find any of
> these tables to have column like "servernum_id" or similar, so my guess is
> that SM is marking these logs somehow, or maybe there is some limitation in
> the Data protector integration.
>
> Regards,
> Lyubomir
>
> 2015-11-16 23:48 GMT+02:00 BRYCE STENBERG <bryce@hrnz.co.nz>:
>
> > Some years ago when we use to still back up to tape, in the batch file that
> > started the full backup, we would first rename
> > '%informixdir%\\\\etc\\\\log_full.bat' to something else so that no triggered log
> > backups could happen during the full backup (I think it gets called by the
> > alarmprogram.bat). Once the full backup completed we'd just rename it back
> > to
> > log_full.bat and the next full log would get all the logs again.
> > Would that work in your situation?
> >
> > Regards,
> > Bryce Stenberg.
> >
> >
> >
> >
>
*******************************************************************************
> > Forum Note: Use "Reply" to post a response in the discussion forum.
> >
> >
>
> --001a11430a368e14ab0524b78eb2
>
>
>
*******************************************************************************
> Forum Note: Use "Reply" to post a response in the discussion forum.
>