RSS setup question using disk clone technology
Chuan Lu set up an RSS from storage snapshots taken without onmode -c block and got recovery errors. Andreas Legner and Art Kagel said the copy was inconsistent and the block and unblock are required. Kirit Rana suggested ifxclone or ontape. David Williams said a consistent group snapshot can work. Davorin Kremenjas urged using the documented blocking checkpoint.
Auto-generated by Claude from the posts below — may be imperfect; read the full thread.
|
view attached (2)
HI,
I want to setup Informix RSS instance using external backup technology (EMC srdf, snapshot and LVM technology), but before doing the external backup, I don't run "onmode -c block".
When the rss node do the logical recovery, it report below errors and panic.
===============================================================
19:02:40 IBM Informix Dynamic Server Started.
19:02:42 Requested shared memory segment size rounded from 67108428KB to 67108432KB
19:02:42 Dynamically allocated new virtual shared memory segment (size 67108432KB)
19:02:42 Memory sizes:resident:67108864 KB, virtual:134217296 KB, no SHMTOTAL limit
19:02:43 Requested shared memory segment size rounded from 67108428KB to 67108432KB
19:02:43 Dynamically allocated new virtual shared memory segment (size 67108432KB)
19:02:43 Memory sizes:resident:67108864 KB, virtual:201325728 KB, no SHMTOTAL limit
19:02:44 Requested shared memory segment size rounded from 67108428KB to 67108432KB
19:02:44 Dynamically allocated new virtual shared memory segment (size 67108432KB)
19:02:44 Memory sizes:resident:67108864 KB, virtual:268434160 KB, no SHMTOTAL limit
19:02:45 Dynamically allocated new virtual shared memory segment (size 50553408KB)
19:02:45 Memory sizes:resident:67108864 KB, virtual:318987568 KB, no SHMTOTAL limit
19:02:46 Requested shared memory segment size rounded from 67108408KB to 67108416KB
19:02:46 Dynamically allocated new virtual shared memory segment (size 67108416KB)
19:02:46 Memory sizes:resident:67108864 KB, virtual:386095984 KB, no SHMTOTAL limit
19:02:47 Requested shared memory segment size rounded from 67108408KB to 67108416KB
19:02:47 Dynamically allocated new virtual shared memory segment (size 67108416KB)
19:02:47 Memory sizes:resident:67108864 KB, virtual:453204400 KB, no SHMTOTAL limit
19:02:48 Requested shared memory segment size rounded from 67108408KB to 67108416KB
19:02:48 Dynamically allocated new virtual shared memory segment (size 67108416KB)
19:02:48 Memory sizes:resident:67108864 KB, virtual:520312816 KB, no SHMTOTAL limit
19:02:49 Dynamically allocated new virtual shared memory segment (size 8192000KB)
19:02:49 Memory sizes:resident:67108864 KB, virtual:528504816 KB, no SHMTOTAL limit
19:02:49 Requested shared memory segment size rounded from 20298676KB to 20298688KB
19:02:49 Dynamically allocated new virtual shared memory segment (size 20298688KB)
19:02:49 Memory sizes:resident:67108864 KB, virtual:548803504 KB, no SHMTOTAL limit
19:02:50 VP pid=7275256 priority fixed at 60
Tue Jul 5 19:03:07 2022
19:03:07 Warning: ONCONFIG dump directory (DUMPDIR) '/informix/tmp' has insecure permissions
19:03:07 Event alarms enabled. ALARMPROG = '/informix/etc/alarmprogram.sh'
19:03:07 Booting Language <c> from module <>
19:03:07 Loading Module <CNULL>
19:03:07 Booting Language <builtin> from module <>
19:03:07 Loading Module <BUILTINNULL>
19:03:19 Trusted host cache successfully built:/etc/hosts.equiv.
19:03:19 DR: DRAUTO is 0 (Off)
19:03:19 DR: ENCRYPT_HDR is 0 (HDR encryption Disabled)
19:03:19 AIX MP latch code enabled
19:03:19 Fast poll pollset enabled.
19:03:21 IBM Informix Dynamic Server Version 11.70.FC8W1XV Software Serial Number AAA#B000000
19:03:23 IBM Informix Dynamic Server Initialized -- Shared Memory Initialized.
19:03:23 Started 1 B-tree scanners.
19:03:23 B-tree scanner threshold set at 5000.
19:03:23 B-tree scanner range scan size set to -1.
19:03:23 B-tree scanner ALICE mode set to 6.
19:03:23 B-tree scanner index compression level set to med.
19:03:23 DR: Reservation of the last logical log for log backup turned on
19:03:23 Data replication type and state information reset. To start DR, use
the 'onmode -d' command and wait for the pair to be operational,
before shutting down the database server
19:03:23 Dataskip is now OFF for all dbspaces
19:03:23 Restartable Restore has been ENABLED
19:03:23 Recovery Mode
19:03:25 The external backup for the root chunk (chunk number 1) is not valid.
19:06:13 DR: Reservation of the last logical log for log backup turned off
19:06:14 DR: new type = RSS
19:06:14 RSS Server sv_sh_rmt - state is now connected
19:06:14 Logical Recovery Started.
19:06:14 113 recovery worker threads will be started.
19:06:14 Start Logical Recovery - Start Log 135661, End Log ?
19:06:14 Starting Log Position - 135661 0x3c245d24
19:06:14 Clearing the physical and logical logs has started
19:06:15 Defragmenter cleaner thread now running
19:06:15 Defragmenter cleaner thread cleaned:0 partitions
19:10:01 Cleared 63581 MB of the physical and logical logs in 227 seconds
19:10:02 B-tree scanners disabled.
19:10:03 DR: RSS secondary server operational
19:52:11 Started processing open transactions on secondary during startup
19:52:11 Finished processing open transactions on secondary during startup.
19:52:11 Assert Warning: Error during recovery left index inconsistent.
19:52:11 Assert Warning: Error during recovery left index inconsistent.
19:52:11 Assert Warning: Error during recovery left index inconsistent.
19:52:11 Assert Warning: Error during recovery left index inconsistent.
19:52:11 IBM Informix Dynamic Server Version 11.70.FC8W1XV
19:52:11 Who: Session(36, informix@HXRSSFZ, 0, 700008c70c696e0)
Thread(261, xchg_1.23, 700008c70c33b68, 9)
File: rskey.c Line: 1648
19:52:11 IBM Informix Dynamic Server Version 11.70.FC8W1XV
19:52:11 Who: Session(36, informix@HXRSSFZ, 0, 700008c70c696e0)
Thread(316, xchg_1.78, 700008c70c502e0, 34)
File: rskey.c Line: 1648
19:52:11 IBM Informix Dynamic Server Version 11.70.FC8W1XV
19:52:11 Who: Session(36, informix@HXRSSFZ, 0, 700008c70c69a18)
Thread(318, xchg_1.80, 700008c70c51370, 21)
File: rskey.c Line: 1648
19:52:11 IBM Informix Dynamic Server Version 11.70.FC8W1XV
19:52:11 Who: Session(36, informix@HXRSSFZ, 0, 700008c70c6aa30)
Thread(259, xchg_1.21, 700008c70c32ad8, 8)
File: rskey.c Line: 1648
19:52:11 Results: Index 'lsk:"cib".qszhlsb#qszhlsbsy1' is now unusable
19:52:11 Results: Index 'lsk:"cib".dqsmx#dqsmxsy' is now unusable
19:52:11 Results: Index 'lsk:"cib".jylssxb#jylssxbsy' is now unusable
19:52:11 Results: Index 'lsk:"cib".jywlxxwj#jywlxxwjsy1' is now unusable
19:52:11 Action: Run 'oncheck -cI lsk:"cib".qszhlsb#qszhlsbsy1'
19:52:11 Action: Run 'oncheck -cI lsk:"cib".dqsmx#dqsmxsy'
19:52:11 Action: Run 'oncheck -cI lsk:"cib".jylssxb#jylssxbsy'
19:52:11 Action: Run 'oncheck -cI lsk:"cib".jywlxxwj#jywlxxwjsy1'
19:52:11 stack trace for pid 6553638 written to /informix/tmp/af.4ed25eb
19:52:11 stack trace for pid 7012742 written to /informix/tmp/af.4eb25eb
19:52:11 stack trace for pid 5964208 written to /informix/tmp/af.52425eb
19:52:11 stack trace for pid 6554258 written to /informix/tmp/af.52625eb
19:52:11 See Also: /informix/tmp/af.4ed25eb
19:52:11 See Also: /informix/tmp/af.4eb25eb
19:52:11 See Also: /informix/tmp/af.52625eb
19:52:11 See Also: /informix/tmp/af.52425eb
19:52:12 Assert Warning: Error during recovery left index inconsistent.
19:52:12 Error during recovery left index inconsistent.
19:52:12 IBM Informix Dynamic Server Version 11.70.FC8W1XV
19:52:12 Who: Session(36, informix@HXRSSFZ, 0, 700008c70c6d408)
Thread(283, xchg_1.45, 700008c70c3f198, 10)
File: rskey.c Line: 1648
19:52:12 Error during recovery left index inconsistent.
19:52:12 Error during recovery left index inconsistent.
19:52:12 Error during recovery left index inconsistent.
19:52:12 Results: Index 'zfk:"cib".sszfdjb#sszfdjbsy3' is now unusable
19:52:12 Action: Run 'oncheck -cI zfk:"cib".sszfdjb#sszfdjbsy3'
19:52:12 stack trace for pid 6750434 written to /informix/tmp/af.50325ec
19:52:13 Assert Warning: Error during recovery left index inconsistent.
19:52:13 IBM Informix Dynamic Server Version 11.70.FC8W1XV
19:52:13 Who: Session(36, informix@HXRSSFZ, 0, 700008c70c6aa30)
Thread(260, xchg_1.22, 700008c70c33320, 11)
File: rskey.c Line: 1648
19:52:13 Assert Warning: Error during recovery left index inconsistent.
19:52:13 See Also: /informix/tmp/af.50325ec
19:52:13 Assert Warning: Error during recovery left index inconsistent.
19:52:13 Assert Warning: Error during recovery left index inconsistent.
19:52:13 Results: Index 'lsk:"cib".jywlxxwj#jywlxxwjsy2' is now unusable
19:52:13 IBM Informix Dynamic Server Version 11.70.FC8W1XV
19:52:13 Who: Session(36, informix@HXRSSFZ, 0, 700008c70c6aa30)
Thread(258, xchg_1.20, 700008c70c32290, 9)
File: rskey.c Line: 1648
19:52:13 IBM Informix Dynamic Server Version 11.70.FC8W1XV
19:52:13 Who: Session(36, informix@HXRSSFZ, 0, 700008c70c696e0)
Thread(262, xchg_1.24, 700008c70c343b0, 12)
File: rskey.c Line: 1648
19:52:13 IBM Informix Dynamic Server Version 11.70.FC8W1XV
19:52:13 Who: Session(36, informix@HXRSSFZ, 0, 700008c70c6ddb0)
Thread(319, xchg_1.81, 700008c70c51bb8, 34)
File: rskey.c Line: 1648
19:52:13 Action: Run 'oncheck -cI lsk:"cib".jywlxxwj#jywlxxwjsy2'
19:52:13 Results: Index 'lsk:"cib".jywlxxwj#jywlxxwjsy' is now unusable
19:52:13 Results: Index 'lsk:"cib".qszhlsb#qszhlsbsy2' is now unusable
19:52:13 Results: Index 'lsk:"cib".jylssxb#jylssxbsy1' is now unusable
19:52:13 stack trace for pid 7537512 written to /informix/tmp/af.4ec25ec
19:52:13 Action: Run 'oncheck -cI lsk:"cib".jywlxxwj#jywlxxwjsy'
19:52:13 Action: Run 'oncheck -cI lsk:"cib".qszhlsb#qszhlsbsy2'
19:52:13 Action: Run 'oncheck -cI lsk:"cib".jylssxb#jylssxbsy1'
19:52:13 stack trace for pid 6553638 written to /informix/tmp/af.4ea25ec
19:52:13 stack trace for pid 7340630 written to /informix/tmp/af.4ee25ec
19:52:13 stack trace for pid 5964208 written to /informix/tmp/af.52725ec
19:52:13 See Also: /informix/tmp/af.4ec25ec
19:52:13 See Also: /informix/tmp/af.4ea25ec
19:52:13 See Also: /informix/tmp/af.4ee25ec
19:52:13 See Also: /informix/tmp/af.52725ec
19:52:13 Error during recovery left index inconsistent.
19:52:13 Error during recovery left index inconsistent.
19:52:13 Error during recovery left index inconsistent.
19:52:13 Error during recovery left index inconsistent.
19:52:13 Error during recovery left index inconsistent.
19:52:14 Rollforward of log record failed. iserrno = 126
19:52:14 Log Record: log = 135661, pos = 0x3c251974, type = OLDRSAM:HINSERT(40), trans = 1204
19:52:14 Assert Failed: Logical log replay error.
19:52:14 IBM Informix Dynamic Server Version 11.70.FC8W1XV
19:52:14 Who: Session(36, informix@HXRSSFZ, 0, 700008c70c6ba48)
Thread(264, xchg_1.26, 700008c70c35440, 24)
File: rsprecvr.c Line: 7465
19:52:14 Results: The secondary server cannot continue.
19:52:14 Action: Reestablish the secondary server.
19:52:14 stack trace for pid 6488196 written to /informix/tmp/af.4f025eb
19:52:14 Rollforward of log record failed. iserrno = 126
19:52:14 Log Record: log = 135661, pos = 0x3c2492f8, type = OLDRSAM:HINSERT(40), trans = 6840
19:52:14 Rollforward of log record failed. iserrno = 126
19:52:14 Log Record: log = 135661, pos = 0x3c247338, type = OLDRSAM:HINSERT(40), trans = 6840
19:52:14 Assert Failed: Logical log replay error.
19:52:14 IBM Informix Dynamic Server Version 11.70.FC8W1XV
19:52:14 Who: Session(36, informix@HXRSSFZ, 0, 700008c70c696e0)
Thread(268, xchg_1.30, 700008c70c37560, 31)
File: rsprecvr.c Line: 7465
19:52:14 Rollforward of log record failed. iserrno = 126
19:52:14 Log Record: log = 135661, pos = 0x3c247110, type = OLDRSAM:HINSERT(40), trans = 940
19:52:14 Assert Failed: Logical log replay error.
19:52:14 Results: The secondary server cannot continue.
19:52:14 IBM Informix Dynamic Server Version 11.70.FC8W1XV
19:52:14 Who: Session(36, informix@HXRSSFZ, 0, 700008c70c696e0)
Thread(295, xchg_1.57, 700008c70c454f8, 14)
File: rsprecvr.c Line: 7465
19:52:14 Action: Reestablish the secondary server.
19:52:14 Results: The secondary server cannot continue.
19:52:14 stack trace for pid 8323592 written to /informix/tmp/af.4f425eb
19:52:14 Assert Failed: Logical log replay error.
19:52:14 Action: Reestablish the secondary server.
19:52:14 IBM Informix Dynamic Server Version 11.70.FC8W1XV
19:52:14 Who: Session(36, informix@HXRSSFZ, 0, 700008c70c6aa30)
Thread(294, xchg_1.56, 700008c70c44cb0, 1)
File: rsprecvr.c Line: 7465
19:52:14 stack trace for pid 6030120 written to /informix/tmp/af.50f25eb
19:52:14 Results: The secondary server cannot continue.
19:52:14 Action: Reestablish the secondary server.
19:52:14 Rollforward of log record failed. iserrno = 126
19:52:14 Log Record: log = 135661, pos = 0x3c2487b4, type = OLDRSAM:HINSERT(40), trans = 1054
19:52:14 stack trace for pid 7275256 written to /informix/tmp/af.50e25eb
19:52:14 Assert Failed: Logical log replay error.
19:52:14 Rollforward of log record failed. iserrno = 126
19:52:14 Log Record: log = 135661, pos = 0x3c2474e8, type = OLDRSAM:HINSERT(40), trans = 7891
19:52:14 IBM Informix Dynamic Server Version 11.70.FC8W1XV
19:52:14 Who: Session(36, informix@HXRSSFZ, 0, 700008c70c69a18)
Thread(289, xchg_1.51, 700008c70c42348, 35)
File: rsprecvr.c Line: 7465
19:52:14 Results: The secondary server cannot continue.
19:52:14 Assert Failed: Logical log replay error.
19:52:14 Action: Reestablish the secondary server.
19:52:14 IBM Informix Dynamic Server Version 11.70.FC8W1XV
19:52:14 Who: Session(36, informix@HXRSSFZ, 0, 700008c70c6ca60)
Thread(272, xchg_1.34, 700008c70c39680, 8)
File: rsprecvr.c Line: 7465
19:52:14 stack trace for pid 5308958 written to /informix/tmp/af.50925eb
19:52:14 Results: The secondary server cannot continue.
19:52:14 Action: Reestablish the secondary server.
19:52:14 stack trace for pid 7012742 written to /informix/tmp/af.4f825ec
19:52:14 See Also: /informix/tmp/af.4f025eb
19:52:14 See Also: /informix/tmp/af.4f425eb
19:52:14 See Also: /informix/tmp/af.50f25eb
19:52:14 See Also: /informix/tmp/af.50e25eb
19:52:14 See Also: /informix/tmp/af.50925eb
19:52:14 See Also: /informix/tmp/af.4f825ec
19:52:14 Rollforward of log record failed. iserrno = 105
19:52:14 Log Record: log = 135661, pos = 0x3c248a74, type = OLDRSAM:ADDITEM(28), trans = 1054
19:52:14 Rollforward of log record failed. iserrno = 105
19:52:14 Log Record: log = 135661, pos = 0x3c248a74, type = OLDRSAM:ADDITEM(28), trans = 1054
19:52:14 Rollforward of log record failed. iserrno = 105
19:52:14 Log Record: log = 135661, pos = 0x3c247898, type = OLDRSAM:ADDITEM(28), trans = 940
19:52:14 Rollforward of log record failed. iserrno = 105
19:52:14 Log Record: log = 135661, pos = 0x3c247898, type = OLDRSAM:ADDITEM(28), trans = 940
19:52:14 Rollforward of log record failed. iserrno = 105
19:52:14 Log Record: log = 135661, pos = 0x3c247688, type = OLDRSAM:ADDITEM(28), trans = 6840
19:52:14 Rollforward of log record failed. iserrno = 105
19:52:14 Log Record: log = 135661, pos = 0x3c247688, type = OLDRSAM:ADDITEM(28), trans = 6840
19:52:14 Rollforward of log record failed. iserrno = 105
19:52:14 Log Record: log = 135661, pos = 0x3c2493f0, type = OLDRSAM:ADDITEM(28), trans = 6840
19:52:14 Rollforward of log record failed. iserrno = 105
19:52:14 Log Record: log = 135661, pos = 0x3c2493f0, type = OLDRSAM:ADDITEM(28), trans = 6840
19:52:14 Rollforward of log record failed. iserrno = 105
19:52:14 Log Record: log = 135661, pos = 0x3c2490bc, type = OLDRSAM:ADDITEM(28), trans = 6756
19:52:14 Rollforward of log record failed. iserrno = 105
19:52:14 Log Record: log = 135661, pos = 0x3c2490bc, type = OLDRSAM:ADDITEM(28), trans = 6756
19:52:14 Rollforward of log record failed. iserrno = 105
19:52:14 Log Record: log = 135661, pos = 0x3c247fa0, type = OLDRSAM:ADDITEM(28), trans = 6840
19:52:14 Rollforward of log record failed. iserrno = 105
19:52:14 Log Record: log = 135661, pos = 0x3c247fa0, type = OLDRSAM:ADDITEM(28), trans = 6840
19:52:14 Rollforward of log record failed. iserrno = 105
19:52:14 Log Record: log = 135661, pos = 0x3c2485f4, type = OLDRSAM:ADDITEM(28), trans = 940
19:52:14 Rollforward of log record failed. iserrno = 105
19:52:14 Log Record: log = 135661, pos = 0x3c2485f4, type = OLDRSAM:ADDITEM(28), trans = 940
19:52:14 Rollforward of log record failed. iserrno = 105
19:52:14 Log Record: log = 135661, pos = 0x3c247550, type = OLDRSAM:ADDITEM(28), trans = 940
19:52:14 Rollforward of log record failed. iserrno = 105
19:52:14 Log Record: log = 135661, pos = 0x3c247550, type = OLDRSAM:ADDITEM(28), trans = 940
19:52:14 Rollforward of log record failed. iserrno = 105
19:52:14 Log Record: log = 135661, pos = 0x3c2491a8, type = OLDRSAM:ADDITEM(28), trans = 2566
19:52:14 Rollforward of log record failed. iserrno = 105
19:52:14 Log Record: log = 135661, pos = 0x3c2491a8, type = OLDRSAM:ADDITEM(28), trans = 2566
19:52:33 Thread ID 295 will NOT be suspended because
it is a daemon.
19:52:33 See Also: /informix/tmp/af.50f25eb
19:52:33 rsprecvr.c, line 7465, thread 295, proc id 6030120, Logical log replay error..
19:52:33 Thread ID 264 will NOT be suspended because
it is a daemon.
19:52:33 See Also: /informix/tmp/af.4f025eb
19:52:33 Thread ID 289 will NOT be suspended because
it is a daemon.
19:52:33 Thread ID 272 will NOT be suspended because
it is a daemon.
19:52:33 See Also: /informix/tmp/af.50925eb
19:52:33 See Also: /informix/tmp/af.4f825ec
19:52:33 rsprecvr.c, line 7465, thread 264, proc id 6488196, Logical log replay error..
19:52:33 rsprecvr.c, line 7465, thread 272, proc id 7012742, Logical log replay error..
19:52:33 rsprecvr.c, line 7465, thread 289, proc id 5308958, Logical log replay error..
19:52:33 Thread ID 294 will NOT be suspended because
it is a daemon.
19:52:33 See Also: /informix/tmp/af.50e25eb
19:52:33 Thread ID 268 will NOT be suspended because
it is a daemon.
19:52:33 See Also: /informix/tmp/af.4f425eb
19:52:33 rsprecvr.c, line 7465, thread 294, proc id 7275256, Logical log replay error..
19:52:33 rsprecvr.c, line 7465, thread 268, proc id 8323592, Logical log replay error..
19:52:33 Fatal error in ADM VP at mt.c:14087
19:52:33 Unexpected virtual processor termination, pid = 6030120, exit = 0x100
19:52:33 PANIC: Attempting to bring system down
=============================================
I attach all the af file and online.log. please take a look, thanks for you time.
I suepend because I don't run the 'onmode -c block', lots of dirty pages are still in the buffer pool, so some indexes and table is inconsistent in the disk clone volumns.
For this problem I have two question .
The first is how can I setup RSS successful using external backup without running 'onmode -c block' ? because I can't run the 'onmode -c block' in the production environment, it' more close to customer DR technology,the DR center in another city use EMC SRDF to do the disk copy real time.
The second question is if I run the instance into quiescent mode firstly using external backup, the purpose is letting those tables and index inconsistent, and I found after the instance doing the physical restore and logical log restore, IDS do the checkpoint . this checkpoint record shouldn't exist in the primary instance. Could I run the instance into recovery mode using oninit -PHY and setup the RSS instance successful?
Any suggestion are welcome.
------------------------------
chuan lu
------------------------------
#Informix
Hi Chuan
Since you are running with IDS
11.70
, you have
ifxclone
available to you. It will make life a lot easier. Failing that, you could still opt to use
ontapepiped to
STDIO
. In both cases you can do away with using the external backup technology that you speak of.
------------------------------
Kirit Rana
------------------------------
What you're doing cannot be called external backup and restore without 'onmode -c block'. Even if your snapshot would block any disk writes during the split/copy, you'd still copy an inconsistent state, for reasons you already stated.
------------------------------
Andreas Legner
------------------------------
Here are your red herrings:
19:03:25 The external backup for the root chunk (chunk number 1) is not valid.
and
19:52:11 Assert Warning: Error during recovery left index inconsistent.
19:52:11 Assert Warning: Error during recovery left index inconsistent.
19:52:11 Assert Warning: Error during recovery left index inconsistent.
19:52:11 Assert Warning: Error during recovery left index inconsistent.
The external backup was not consistent. The root chunk's consistency page does not show a final, archive, checkpoint resulting in some index nodes being inconsistent.
You MUST perform and "onmode -c block" before the snapshot and "onmode -c unblock" after it completes. Also I'm not sure that you can initiate a secondary from an external archive!
------------------------------
Art S. Kagel, President and Principal Consultant
ASK Database Management Corp.
www.askdbmgt.com
------------------------------
A true external backup, i.e. an 'onmode -c block' consistent copy of your instance's chunks, is as good as (or better even than) a physically restored L 0 backup, so yes, this is definitely one way to instantiate a secondary server.
------------------------------
Andreas Legner
------------------------------
Thanks Andreas. Personally I still trust ontape and onbar over any external backup, but, good to know.
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.
Hi,
You can do this without onmode -c block however...you would have to snapshot all chunks at the same time as a consistent group.
Are you using raw devices or files on a filesystem?
I think this is possible if EITHER:
a) All your chunks are files on the SAME filesystem.and you use an LVM snapshot of the filesystem.
OR
b) In EMC SDRF if there is more than 1 volume you create a volume consistency group.and snapshot the group.
Then you can start the secondary with oninit -PHY and onmode -d RSS.
Regards,
David.
------------------------------
David Williams
------------------------------
HI,
We did the testing again yesterday night.
we still doesn't use the 'onmode -c block' before doing the external backup, because it's a production environment.
We found "oninit" could bring the instance online, and no inconsistent errors in online.log;
After that we recover the data from the snapshot, and try to setup the RSS with "ontape -p -e " and "onmode -d RSS xxx", we see the inconsistent errors again. Why? is it informix defect in this version?
I compare the online.log , and found before RSS setup IDS will clear the physical log and logical log, What's the RSS instance doing? it's different with the physical recovery and logical recovery which oninit do.
Lu chuan
------------------------------
chuan lu
------------------------------
If done properly (all chunks at a time), a snapshot between two checkpoints would provide you a copied state on disk that is very similar to that after a server crash, so fast recovery could take care of this and get this (copy of your) instance up, as a new standalone server. What such snapshot does not provide is a defined starting point for a logical restore which essentially is what a secondary server does. ------------------------------ Andreas Legner ------------------------------
Hi chuan, why even bother? The docs clearly state it and others have already commented - you should use the blocking checkpoint. Not sure what the production environment arguments stands for - we've done it literally thousands of times (every night for many years, scripted/unattended from the cron job) on large instances (tens of terabytes), the most time it ever took was around 12 seconds. The trick is to have the underlying storage technology keeping up the secondary group of LUNs (which will become the RSS chunks) very close to the primary LUN consistency group, with as little delta in between them as possible and then breaking/fracturing the mirror at a quiet time where 1) it doesn't take too long to do it as there are very little new writes advancing the primary LUN group position and 2) the impact to the business is minimal due to the same reason. Can't remember ever having a single major issue when performing the procedure the right/documented way. ------------------------------ Davorin Kremenjas ------------------------------