Re: Enterprise Replication question
- From: Madison Pruet <mpruet1@xxxxxxxxxxx>
- Date: Thu, 02 Aug 2007 03:10:16 GMT
Jarrod Teale wrote:
The output looks fine (I've set it up for five minutes) as below but the
transactions are still coming back VERY fast (each second or two)
I've included date stamp and current record data from the target table
on the central server - as you can see it's pretty much immediate up to
date (the servers are time synced.)
I wouldn't do timed based replication. I put in a whole bunch of warning messages when it is used. ;-)... The reason is that when using timed based replication, the transaction is split into multiple transmission units and you simply can't support things such as referential integrity when doing that.
A possible solution for you which would allow you to maintain the transactional unit - and to control the transmissions would be to use "cdr disconnect" and "cdr connect" periodically. This would cause the network connection to be dropped between the servers and then later reestablished. After running cdr connect, you could wait for a while then issue a disconnect.
As an example, suppose that you had the following deamon...
while [ 1 ];do
cdr connect -c <source> <target>
sleep 60
cdr disconnect -c <source> <target>
sleep 600
done
This would allow traffic for a min., then would disconnect and remain disconnected for 10 min. Then reconnect, etc....
M.P.
.
I want to look at the -e 10 option to help with compression
I have CDR_NIFCOMPRESS set to 9 and I want to see what difference in
network traffic occurs when I batch the transactions
I operate in a constrained network environment where I don't have full
control of my WAN links - The less I use them the less I get yelled at.
If I batch the data into 3-10 minute blocks, I'm hoping it will bring my
daily volume of data transferred between servers down.
I can happily live with the data being this old in the central system(it
updates on the source system every few seconds)
[informix@sirrp71(10.71.64.130) ~]$ onstat -g rep
IBM Informix Dynamic Server Version 10.00.UC3 -- On-Line -- Up 14
days 03:18:50 -- 144028 Kbytes
Time based replicates
Name Frequency Last Next
=======================================================================
stan_set Every 5 min Aug 2 14:38:11 Aug 2 14:43:11
Replset
sirrp71_stan_tags Every 5 min Aug 2 14:38:11 Aug 2 14:43:11
sirrp71_stan_tagsa Every 5 min Aug 2 14:38:11 Aug 2 14:43:11
sirrp71_apcapp_cyp Every 30 min Aug 2 14:30:38 Aug 2 15:00:38
apcappstatus_set Every 30 min Aug 2 14:30:38 Aug 2 15:00:38
Replset
Schedule manager Cb: 0x13fd6e18 State: 0x8100 <CDRINIT,CDRRUNNING>
Event Thread When
============================================
CDRDS CDREvent 00:00:04
CDRRepSched CDRRep 00:04:37
CDRRepSched CDRRep 00:04:37
CDRRepSched CDRRep 00:04:37
CDRRepSched CDRRep 00:22:04
CDRRepSched CDRRep 00:22:04
[sdev@hades ~]$ date
Thu Aug 2 14:40:41 NZST 2007
[sdev@hades ~]$ dbaccess stan <<!
select max(sample_dt) from tagsamples_71
!
Database selected.
(max)
2007-08-02 14:40:42.000
1 row(s) retrieved.
Database closed.
-----Original Message-----
From: Madison Pruet [mailto:mpruet1@xxxxxxxxxxx] Sent: Thursday, 2 August 2007 2:32 p.m.
To: Jarrod Teale
Cc: informix-list@xxxxxxxx
Subject: Re: Enterprise Replication question
Jarrod Teale wrote:Any chance of some assistance with this? I've had a further look and I
can't make the -e option work anywhere in my ER replicates, from starting new ones to changing existing ones or through the use of a replicate set.
Is there something I'm missing here? Do I need to turn this feature on
somewhere - or tell the instance to use it?
I will note that there are a number of other transaction based replicates running between the two servers and a heap of other servers
also sending data back to the hades server (destination). ER doesn't just run all replicates at the fastest of any of those defined doesit?
I just tested this and it seemed to be working ok. What is the output
of onstat -g rep on the source system. (With timed based replication,
it's the source system that manages the transmission of the changes.)
Just as an aside, what is the reason that you are wanting to use the -e
10 option?
Jarrodbeautifully
-----Original Message-----
From: informix-list-bounces@xxxxxxxx
[mailto:informix-list-bounces@xxxxxxxx] On Behalf Of Jarrod Teale
Sent: Monday, 30 July 2007 4:14 p.m.
To: informix-list@xxxxxxxx
Subject: RE: Enterprise Replication question
Madison,
I defined the replicate like this
cdr define repl -C ignore sirrp71_stan_tagsamples \tagsamples_71" \
"R stan@g_hades:sdev.tagsamples_71" "SELECT * FROM
"P stan@g_sirrp71:sdev.tagsamples" "SELECT * FROM tagsamples"Made this adjustment
cdr modify repl --ignoredel y sirrp71_stan_tagsamplesThen made the time adjustment
cdr modify repl -e 10 sirrp71_stan_tagsamplesI've also tried deleting the replicate and recreating with the -e flag
in the define step but still see transactions rolling in every few
seconds
Thanks
Jarrod
-----Original Message-----
From: Madison Pruet [mailto:mpruet1@xxxxxxxxxxx]
Sent: Monday, 30 July 2007 4:09 p.m.
To: Jarrod Teale
Cc: informix-list@xxxxxxxx
Subject: Re: Enterprise Replication question
Jarrod Teale wrote:Hi,
Running 10.00.UC3 -> 10.00.UC3 replication, both on RedHat EL 4
I set up a one-way replicate to transfer data and it works(Go Informix) but I forgot to set the frequency option in the replicate.Please post the command you used to define the replicate and also the
'cdr modify repl' command that you issued.
M.P.
I've now set it to the right thing, using cdr modify replicate but it
----------------------------------------------------------------------doesn't seem to have taken effect. (You can see the 10 minute frequency
below)
The records are still arriving at the central server at the end of each transaction, not caching for 10 minutes before being sent
Any ideas?
Jarrod Teale
Fonterra New Zealand.
[informix@sirrp71(10.71.64.130) ~]$ cdr list repl sirrp71_stan_tagsamples
DEFINED REPLICATES ATTRIBUTES
------------------------------
REPLICATE: sirrp71_stan_tagsamples
STATE: Active ON:sirrp71
CONFLICT: Ignore
FREQUENCY: every 00:10
QUEUE SIZE: 0
PARTICIPANT: stan:sdev.tagsamples
OPTIONS: transaction,fullrow
[informix@sirrp71(10.71.64.130) ~]$ cdr list repl brief sirrp71_stan_tagsamples
REPLICATE TABLE SELECT
any------ sirrp71_stan_tagsamples stan@g_sirrp71:sdev.tagsamplestagsamples
<mailto:stan@g_sirrp71:sdev.tagsamples> SELECT * FROM
sirrp71_stan_tagsamples db@g_hades:owner.table <mailto:db@g_hades:owner.table> select * from tablethis email.
DISCLAIMER:
This email contains confidential information and may be legally privileged. If you are not the intended recipient or have received this email in error, please notify the sender immediately and destroyYou may not use, disclose or copy this email or its attachments inway.thisAny opinions expressed in this email are those of the author and are not necessarily those of the Fonterra Co-operative Group.
http://www.fonterra.com/
DISCLAIMER:
This email contains confidential information and may be legally
privileged. If you are not the intended recipient or have received
email in error, please notify the sender immediately and destroy thisnot
email.
You may not use, disclose or copy this email or its attachments in any
way.
Any opinions expressed in this email are those of the author and are
necessarily those of the Fonterra Co-operative Group.privileged. If you are not the intended recipient or have received this
http://www.fonterra.com/
_______________________________________________
Informix-list mailing list
Informix-list@xxxxxxxx
http://www.iiug.org/mailman/listinfo/informix-list
DISCLAIMER:
This email contains confidential information and may be legally
email in error, please notify the sender immediately and destroy this
email.
You may not use, disclose or copy this email or its attachments in anyway.Any opinions expressed in this email are those of the author and arenot necessarily those of the Fonterra Co-operative Group.http://www.fonterra.com/
DISCLAIMER:
This email contains confidential information and may be legally privileged. If you are not the intended recipient or have received this email in error, please notify the sender immediately and destroy this email.
You may not use, disclose or copy this email or its attachments in any way.
Any opinions expressed in this email are those of the author and are not necessarily those of the Fonterra Co-operative Group.
http://www.fonterra.com/
- References:
- RE: Enterprise Replication question
- From: Jarrod Teale
- RE: Enterprise Replication question
- Prev by Date: RE: Enterprise Replication question
- Next by Date: System or internal errorDSRA0010E: SQL State = IX000, Error Code = -79716
- Previous by thread: RE: Enterprise Replication question
- Next by thread: System or internal errorDSRA0010E: SQL State = IX000, Error Code = -79716
- Index(es):
Relevant Pages
|