Re: archecker redux.....no P word

Hi, There are these issues -- 1) Do not use the -l phys option -- this does not mean to do physical restore only -- it means that you *intend* to do logical restore, but want to do the physical part and the logical part in separate commands. 2) Grep for ERROR from the archecker log -- it will tell you if there's a reason why it's skipping the restore of the table -- there may be problems with creating the table to be restored (if the table hasn't been created already) or some other issue with access to the table. For instance, you might be having an issue with DB_LOCALE -- but you'll be able to tell this if you grep for ERROR from the AC_MSGPATH -- errors are not always reported on the stdout/stderr. archecker assumes DB_LOCALE of en_us.8859-1 unless you set it to something else before running archecker. You should set the environment DB_LOCALE to the DB_LOCALE of the database before running archecker. 3) Run archecker -DX to clean up all previous archecker working files and tables before running a new archecker table-level restore command. Or, you can add the -d option to archecker table-level restore command to automatically remove all previous working files and tables before running the current command. Additional Details: Errors are not always reported on the stdout/stderr -- you need to grep for "ERROR" or "WARN" from the log file. One warning you should see is that the filter is being disabled -- this is because you're using the -lphys option which indicates your *intention* to do logical recovery by following with additional archecker commands with options for -l stage,apply . You should *not* use -l phys if you have the "restore to current with no log;" in your schema command file. Using "-l phys" and using "restore to current with no log;" are not equivalent -- they mean different things. To use filters, you should use "restore to current with no log;" only. So, your archecker command should just be -- archecker -Xtvs Side note: ---------------------- I also like to add the -d option to make sure that archecker cleans up all previous working tables and files from previous archecker runs before running the current archecker command -- this isn't necessary if you remember to run "archecker -DX" to clean up after running archecker commands, but it doesn't hurt to use the -d option always. Also, you do not need the -X option if you use the -f option to specify your schema command file. Example: archecker -tdvs -f <schema command file> ------------------------ --Hyun-Ju ===================================== From: Tam O'Shanter <tam@oshanter.com> To: informix-list@iiug.org Sent: Wednesday, November 15, 2006 7:02:54 AM Subject: Re: archecker redux.....no P word Yes. It does. Any other suggestions? Tam. "Superboer" <superboer7@t-online.de> wrote in message news:1163588676.389674.158600@i42g2000cwa.googlegroups.com... >i would restore the backup to a test box and check if it is there > including > checking if > >> Select * from book_history>> >> Where book_no > 10131825; > > returns any rows. > > > Superboer. > > > Tam O'Shanter schreef: > >> Okay, >> >> Here is a copy of the command used, the schema file used and the outup >> when >> I last ran archecker. >> >> The frustrating bit is this: >> >> Table extraction commands 1 >> >> Tables found on archive 0 >> >> Tables not located on archive 1 >> >> I know the table is on this backup. >> >> Any thoughts? >> >> This is really strange and quite frustrating. >> >> I'm hoping it's something stupid I'm doing. >> >> ---------------------------------- >> Command: >> date >> >> archecker -Xtvs -lphys >> >> date >> >> ----------------------------------- >> Schema File: >> >> database my_qa_db; >> create table book_history (>> >> book_no INT not null, >> >> crea_date_time CHAR(19) not null, >> >> hist_seqn_no SMALLINT not null, >> >> history_code CHAR(3) not null, >> >> tran_type CHAR(2) not null, >> >> item_seqn_no SMALLINT not null, >> >> item_text VARCHAR(255), >> >> received_from VARCHAR(60), >> >> processing_flg CHAR(1), >> >> duty_code CHAR(5) not null, >> >> branch_code CHAR(12) not null, >> >> crea_user_code CHAR(5) not null, >> >> crea_dest_id CHAR(8) not null >> >> ) in dbs4 ; >> >> insert into book_history>> >> Select * from book_history>> >> Where book_no > 10131825; >> >> restore to current with no log; >> >> >> ------------------------------------------- >> Output: >> >> >> >> Tue Nov 14 16:50:00 EST 2006 >> >> IBM Informix Dynamic Server Version 10.00.UC4 >> >> Program Name: archecker >> >> Version: 8.0 >> >> Released: 2005-11-18 22:22:21 >> >> CSDK: IBM Informix CSDK Version 2.90 >> >> ESQL: IBM Informix-ESQL Version 2.90.UC1 >> >> Compiled: 11/18/05 22:23 on Linux 2.4.21-27.0.2.ELsmp #1 SMP Wed >> Jan >> 12 2 >> >> 3:35:44 EST 2005 >> >> >> >> Debug level set to LOW >> >> Debug Components set to 0xFFFFC >> >> Debugging for component parser - enabled >> >> Debugging for component sql - enabled >> >> Debugging for component sblob - enabled >> >> Debugging for component lowlevel - enabled >> >> Debugging for component partition - enabled >> >> Debugging for component acu - enabled >> >> Debugging for component chunk - enabled >> >> Debugging for component page - enabled >> >> Debugging for component ipa - enabled >> >> Debugging for component eval - enabled >> >> Debugging for component onbar - enabled >> >> Debugging for component ontape - enabled >> >> Debugging for component log - enabled >> >> Debugging for component apply - enabled >> >> Debugging for component stager - enabled >> >> Debugging for component other - enabled >> >> AC_STORAGE /tmp >> >> AC_MSGPATH /opt/informix/my_web/ac_chk.log >> >> AC_VERBOSE on >> >> AC_TAPEDEV /dev/st0 >> >> AC_TAPEBLOCK 512 KB >> >> AC_LTAPEDEV /dev/null >> >> AC_LTAPEBLOCK 256 KB >> >> Command line is overriding command file RESTORE option >> >> Extracting table my_qa_db:book_history into my_qa_db:book_history >> >> >> >> Please put in Phys Tape 1. >> >> Type <return> or 0 to end: >> >> >> >> Tape type: Archive Backup Tape >> >> OnLine version: IBM Informix Dynamic Server Version 10.00.UC4 >> >> Archive date: Tue Nov 14 00:01:02 2006 >> >> Archive level: 0 >> >> Tape blocksize: 524288@@NL

Wow. That seems to work!! Thank you so much. The docs on this are rather misleading, but your suggestions seemed to work. Tested it on a small table with 1.2M rows, now have a test running on a 280 M rows table so my fingers are crossed. Thanks, Tam. "Hyun-Ju Vega" <vegah@us.ibm.com> wrote in message news:mailman.276.1163630536.29126.informix-list@iiug.org... Hi, There are these issues -- 1) Do not use the -l phys option -- this does not mean to do physical restore only -- it means that you *intend* to do logical restore, but want to do the physical part and the logical part in separate commands. 2) Grep for ERROR from the archecker log -- it will tell you if there's a reason why it's skipping the restore of the table -- there may be problems with creating the table to be restored (if the table hasn't been created already) or some other issue with access to the table. For instance, you might be having an issue with DB_LOCALE -- but you'll be able to tell this if you grep for ERROR from the AC_MSGPATH -- errors are not always reported on the stdout/stderr. archecker assumes DB_LOCALE of en_us.8859-1 unless you set it to something else before running archecker. You should set the environment DB_LOCALE to the DB_LOCALE of the database before running archecker. 3) Run archecker -DX to clean up all previous archecker working files and tables before running a new archecker table-level restore command. Or, you can add the -d option to archecker table-level restore command to automatically remove all previous working files and tables before running the current command. Additional Details: Errors are not always reported on the stdout/stderr -- you need to grep for "ERROR" or "WARN" from the log file. One warning you should see is that the filter is being disabled -- this is because you're using the -lphys option which indicates your *intention* to do logical recovery by following with additional archecker commands with options for -l stage,apply . You should *not* use -l phys if you have the "restore to current with no log;" in your schema command file. Using "-l phys" and using "restore to current with no log;" are not equivalent -- they mean different things. To use filters, you should use "restore to current with no log;" only. So, your archecker command should just be -- archecker -Xtvs Side note: ---------------------- I also like to add the -d option to make sure that archecker cleans up all previous working tables and files from previous archecker runs before running the current archecker command -- this isn't necessary if you remember to run "archecker -DX" to clean up after running archecker commands, but it doesn't hurt to use the -d option always. Also, you do not need the -X option if you use the -f option to specify your schema command file. Example: archecker -tdvs -f <schema command file> ------------------------ --Hyun-Ju ===================================== From: Tam O'Shanter <tam@oshanter.com> To: informix-list@iiug.org Sent: Wednesday, November 15, 2006 7:02:54 AM Subject: Re: archecker redux.....no P word Yes. It does. Any other suggestions? Tam. "Superboer" <superboer7@t-online.de> wrote in message news:1163588676.389674.158600@i42g2000cwa.googlegroups.com... >i would restore the backup to a test box and check if it is there > including > checking if > >> Select * from book_history>> >> Where book_no > 10131825; > > returns any rows. > > > Superboer. > > > Tam O'Shanter schreef: > >> Okay, >> >> Here is a copy of the command used, the schema file used and the outup >> when >> I last ran archecker. >> >> The frustrating bit is this: >> >> Table extraction commands 1 >> >> Tables found on archive 0 >> >> Tables not located on archive 1 >> >> I know the table is on this backup. >> >> Any thoughts? >> >> This is really strange and quite frustrating. >> >> I'm hoping it's something stupid I'm doing. >> >> ---------------------------------- >> Command: >> date >> >> archecker -Xtvs -lphys >> >> date >> >> ----------------------------------- >> Schema File: >> >> database my_qa_db; >> create table book_history (>> >> book_no INT not null, >> >> crea_date_time CHAR(19) not null, >> >> hist_seqn_no SMALLINT not null, >> >> history_code CHAR(3) not null, >> >> tran_type CHAR(2) not null, >> >> item_seqn_no SMALLINT not null, >> >> item_text VARCHAR(255), >> >> received_from VARCHAR(60), >> >> processing_flg CHAR(1), >> >> duty_code CHAR(5) not null, >> >> branch_code CHAR(12) not null, >> >> crea_user_code CHAR(5) not null, >> >> crea_dest_id CHAR(8) not null >> >> ) in dbs4 ; >> >> insert into book_history>> >> Select * from book_history>> >> Where book_no > 10131825; >> >> restore to current with no log; >> >> >> ------------------------------------------- >> Output: >> >> >> >> Tue Nov 14 16:50:00 EST 2006 >> >> IBM Informix Dynamic Server Version 10.00.UC4 >> >> Program Name: archecker >> >> Version: 8.0 >> >> Released: 2005-11-18 22:22:21 >> >> CSDK: IBM Informix CSDK Version 2.90 >> >> ESQL: IBM Informix-ESQL Version 2.90.UC1 >> >> Compiled: 11/18/05 22:23 on Linux 2.4.21-27.0.2.ELsmp #1 SMP Wed >> Jan >> 12 2 >> >> 3:35:44 EST 2005 >> >> >> >> Debug level set to LOW >> >> Debug Components set to 0xFFFFC >> >> Debugging for component parser - enabled >> >> Debugging for component sql - enabled >> >> Debugging for component sblob - enabled >> >> Debugging for component lowlevel - enabled >> >> Debugging for component partition - enabled >> >> Debugging for component acu - enabled >> >> Debugging for component chunk - enabled >> >> Debugging for component page - enabled >> >> Debugging for component ipa - enabled >> >> Debugging for component eval - enabled >> >> Debugging for component onbar - enabled >> >> Debugging for component ontape - enabled >> >> Debugging for component log - enabled >> >> Debugging for component apply - enabled >> >> Debugging for component stager - enabled >> >> Debugging for component other - enabled >> >> AC_STORAGE /tmp >> >> AC_MSGPATH /opt/informix/my_web/ac_chk.log >> >> AC_VERBOSE on >> >> AC_TAPEDEV /dev/st0 >> >> AC_TAPEBLOCK 512 KB >> >> AC_LTAPEDEV /dev/null >> >> AC_LTAPEBLOCK 256 KB >> >> Command line is overriding command file RESTORE option >> >> Extracting table my_qa_db:book_history into my_qa_db:book_history >> >> >> >> Please put in Phys Tape 1. >> >> Type <return> or 0 to end: >> >> >> >> Tape type: Archive Backup Tape >> >> OnLi