Sybase NNTP forums - End Of Life (EOL)

The NNTP forums from Sybase - forums.sybase.com - are now closed.

All new questions should be directed to the appropriate forum at the SAP Community Network (SCN).

Individual products have links to the respective forums on SCN, or you can go to SCN and search for your product in the search box (upper right corner) to find your specific developer center.

out sequence tran loads

6 posts in Unix Last posting was on 2001-04-04 19:21:39.0Z
George Posted on 2001-03-30 21:55:36.0Z
From: george
Date: Fri, 30 Mar 2001 17:55:36 -0400
Newsgroups: sybase.public.sqlserver.unix
Subject: out sequence tran loads
Message-ID: <3FECC5EF3C28B8540078725C85256A1F.0078728185256A1F@webforums>
Lines: 20
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Path: forums-1-dub!forums-master.sybase.com!forums.sybase.com!webforums.sybase.com!news
Xref: forums-1-dub sybase.public.sqlserver.unix:669
Article PK: 1092760

Ok folks help me solve this one.

I have a db running on 11.5.1 on Solaris 2.6 on a SUN 3500 4 proccessors.
DB size is 34 gb. We dump trans every 15 minutes and reload on a warm
standby at a remote site. Problem is this: every now and then something
causes us to have a non log transaction on the primary db or we may need to
recycle the warm server. Basic problem is that any such event causes my
logs to be out of sequence and I need to do a complete dump and load and
then apply the tran dumps. This takes about 4 hours. We do database dumps
twice at night so tran dumps typically run wee hours to next night. If I
have to to a full load it means applying tran dumbs from 1:15 am to
whenever in the day I am. How can I, even if I need to restart my warm
standby continue to apply tran dumps with out having to do a load database
first?

I have repserver and am in the process of extensive testing. This was up
but because of all the problems (logs filling up too quick) was shelved by
management. Untill we have tested and have it up and running tran dumps are
what we will be doing. So is there a dbcc command or some such that can
force the tran load to my warm standby and resink all the remaining tran
dumps lined up to load?


Dinnyes Ferenc Posted on 2001-04-02 14:49:14.0Z
Message-ID: <3AC8916A.A17A3E08@sybase.hu>
Date: Mon, 02 Apr 2001 16:49:14 +0200
From: Dinnyes Ferenc <dinnyesf@sybase.hu>
Organization: Axis Kft
X-Mailer: Mozilla 4.7 [en] (WinNT; I)
X-Accept-Language: en,hu
MIME-Version: 1.0
Subject: Re: out sequence tran loads
References: <3FECC5EF3C28B8540078725C85256A1F.0078728185256A1F@webforums>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Newsgroups: sybase.public.sqlserver.unix
Lines: 41
NNTP-Posting-Host: gate.axis.hu 194.149.29.6
Path: forums-1-dub!forums-master.sybase.com!forums.sybase.com
Xref: forums-1-dub sybase.public.sqlserver.unix:660
Article PK: 1092749

If I understood correctly your problem, you are doing unlogged
transactions like fast bcp, select into, create index or so..

Try to avoid these kind of trans, if you want to use the log
keeping warm standby database..

You can schedule these kind of trans right before DB dump,
for instance...

george wrote:
>
> Ok folks help me solve this one.
>
> I have a db running on 11.5.1 on Solaris 2.6 on a SUN 3500 4 proccessors.
> DB size is 34 gb. We dump trans every 15 minutes and reload on a warm
> standby at a remote site. Problem is this: every now and then something
> causes us to have a non log transaction on the primary db or we may need to
> recycle the warm server. Basic problem is that any such event causes my
> logs to be out of sequence and I need to do a complete dump and load and
> then apply the tran dumps. This takes about 4 hours. We do database dumps
> twice at night so tran dumps typically run wee hours to next night. If I
> have to to a full load it means applying tran dumbs from 1:15 am to
> whenever in the day I am. How can I, even if I need to restart my warm
> standby continue to apply tran dumps with out having to do a load database
> first?
>
> I have repserver and am in the process of extensive testing. This was up
> but because of all the problems (logs filling up too quick) was shelved by
> management. Untill we have tested and have it up and running tran dumps are
> what we will be doing. So is there a dbcc command or some such that can
> force the tran load to my warm standby and resink all the remaining tran
> dumps lined up to load?


George Posted on 2001-04-02 22:25:51.0Z
From: george
Date: Mon, 2 Apr 2001 18:25:51 -0400
Newsgroups: sybase.public.sqlserver.unix
Subject: Re: out sequence tran loads
Message-ID: <3A4B14C87FA0D428007B376585256A22.0053854D85256A22@webforums>
References: <3FECC5EF3C28B8540078725C85256A1F.0078728185256A1F@webforums> <3AC8916A.A17A3E08@sybase.hu>
Lines: 2
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Path: forums-1-dub!forums-master.sybase.com!forums.sybase.com!webforums.sybase.com!news
Xref: forums-1-dub sybase.public.sqlserver.unix:658
Article PK: 1092748

I know...but in the event that the non log transactions take place then is
there a way to fix it..other than do a full load?


Bret Halford Posted on 2001-04-02 22:49:57.0Z
Message-ID: <3AC90215.82D5F488@sybase.com>
Date: Mon, 02 Apr 2001 16:49:57 -0600
From: Bret Halford <bret@sybase.com>
Organization: Sybase, Inc.
X-Mailer: Mozilla 4.5 [en]C-CCK-MCD (WinNT; I)
X-Accept-Language: en
MIME-Version: 1.0
Subject: Re: out sequence tran loads
References: <3FECC5EF3C28B8540078725C85256A1F.0078728185256A1F@webforums> <3AC8916A.A17A3E08@sybase.hu> <3A4B14C87FA0D428007B376585256A22.0053854D85256A22@webforums>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Newsgroups: sybase.public.sqlserver.unix
Lines: 12
NNTP-Posting-Host: bret-pc.sybase.com 157.133.80.211
Path: forums-1-dub!forums-master.sybase.com!forums.sybase.com
Xref: forums-1-dub sybase.public.sqlserver.unix:657
Article PK: 1092747

No. Although you might be able to greatly improve the time it takes
to do a full backup - have a look at what the combination of QUIESCE DATABASE
(a new feature in ASE 12.0) and EMC's TimeFinder (rapid disk copy system)
can do for you.
http://www.emc.com/products/product_pdfs/wp/timefinder/SybaseASE12.pdf

-bret

george wrote:

> I know...but in the event that the non log transactions take place then is
> there a way to fix it..other than do a full load?


George Posted on 2001-04-04 19:21:39.0Z
From: George
Date: Wed, 4 Apr 2001 15:21:39 -0400
Newsgroups: sybase.public.sqlserver.unix
Subject: Re: out sequence tran loads
Message-ID: <321E498FC4C72280006A5A0C85256A24.0081350A85256A22@webforums>
References: <3FECC5EF3C28B8540078725C85256A1F.0078728185256A1F@webforums> <3AC8916A.A17A3E08@sybase.hu> <3A4B14C87FA0D428007B376585256A22.0053854D85256A22@webforums> <3AC90215.82D5F488@sybase.com>
Lines: 6
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Path: forums-1-dub!forums-master.sybase.com!forums.sybase.com!webforums.sybase.com!news
Xref: forums-1-dub sybase.public.sqlserver.unix:636
Article PK: 1092726

In another situation...If my warmdb needs recycling (perhaps because
another db on the server has been marked suspect and we need to drop it
with dbcc dbrepair) can I continue to load my tran dumps or will the online
recovery mark the log on my warm db and therefore make the yet to be loaded
trandumbs out of sequence? An if this is the case is there a means of
bringing the warm db up (not with bypass recovery) but without marking the
log


George Posted on 2001-04-03 16:30:30.0Z
From: george
Date: Tue, 3 Apr 2001 12:30:30 -0400
Newsgroups: sybase.public.sqlserver.unix
Subject: Re: out sequence tran loads
Message-ID: <59F71B55866D3487005AAEBD85256A23.0081350A85256A22@webforums>
References: <3FECC5EF3C28B8540078725C85256A1F.0078728185256A1F@webforums> <3AC8916A.A17A3E08@sybase.hu> <3A4B14C87FA0D428007B376585256A22.0053854D85256A22@webforums> <3AC90215.82D5F488@sybase.com>
Lines: 49
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Path: forums-1-dub!forums-master.sybase.com!forums.sybase.com!webforums.sybase.com!news
Xref: forums-1-dub sybase.public.sqlserver.unix:647
Article PK: 1092737

On a related note. In an effort to prevent users going into my warm standby
it was set to single user mode. On a couple of occasions my tran loads have
returned an error 6902 as follows:

02:00000:00043:2001/04/03 08:03:36.11 server Error: 6902, Severity: 21,
State:
0
02:00000:00043:2001/04/03 08:03:36.12 server Page timestamp value falls
between
the old and new timestamps from log. Page #=17204088, object id =
1205631388, p
age timestamp=0001 d08a437c. Log: old timestamp=0001 d08a437b, new
timestamp=0
001 d08f6f9d.
02:00000:00043:2001/04/03 08:03:36.12 server Diagnostics for error 6902
while r
edoing log row rid (4938180,3) in database 'finprod' of xact id (0,0) for
spid 4
3 follow.
LOG SCAN DEFINTION:
Database id : 6
Forward scan: starting at (4938180,3)
maximum of 1 log records.

LOG RECORDS:
DELETE (4938180,3) [REPLICATE]
attcnt=1 rno=3 op=5 padlen=2 sessionid = 4938174,8 len=256
odc_stat=0x00
08
(0x0008 (LHSC_REPLICATE))
loh_status: 0x8 (0x00000008 (LHSC_REPLICATE))
objid=1205631388 pageno=17204088 offset=32 status=0x00 cid=0
old ts=0x0001 0xd08a437b new ts=0x0001 0xd08f6f9d


Total number of log records 1
00:00000:00024:2001/04/03 08:05:03.29 server WARNING:
*************************
**
00:00000:00024:2001/04/03 08:05:03.29 server Attempt by user 17 to dump
xact on
db ubjconv with NO_LOG
00:00000:00024:2001/04/03 08:05:03.40 server Attempt by user 17 to dump
xact on
--More--

-------------------------------------
I read the message but could the difference in options between live and my
backup (Backup in single user mode) be a contributing factor?