Slow Controls Time History: Difference between revisions
No edit summary |
No edit summary |
||
Line 11: | Line 11: | ||
0,15,30,45 * * * * /home/epics/history/www/check_get.pl | 0,15,30,45 * * * * /home/epics/history/www/check_get.pl | ||
If started by cron job, log file ''/home/epics/history/log/get_latest_cont.log'' will be produced. | CURRENTLY GOES TO /dev/null !!! If started by cron job, log file ''/home/epics/history/log/get_latest_cont.log'' will be produced. | ||
Database ''epics'' on ''clondb1'' was creates as follwoing: | Database ''epics'' on ''clondb1'' was creates as follwoing: | ||
Line 26: | Line 26: | ||
DUMPING FROM clasweb-bck: | IMPORTANT: DUMPING FROM clasweb-bck: | ||
cd /home | cd /home | ||
Line 32: | Line 32: | ||
mysqldump: Can't get CREATE TABLE for table `hallb_sf_xy560_0_12_1min` (Can't open file: 'hallb_sf_xy560_0_12_1min.MYI'. (errno: 145)) | mysqldump: Can't get CREATE TABLE for table `hallb_sf_xy560_0_12_1min` (Can't open file: 'hallb_sf_xy560_0_12_1min.MYI'. (errno: 145)) | ||
Following command produce error message but runs: | Following command produce error message but runs (took about 33 minutes): | ||
mysqldump -f -l --opt -uroot -pe1tocome epics | gzip > /home/epics.sql.gz | mysqldump -f -l --opt -uroot -pe1tocome epics | gzip > /home/epics.sql.gz | ||
Line 46: | Line 46: | ||
gunzip epics.sql.gz | gunzip epics.sql.gz | ||
mysql -u root -p epics < epics.sql | mysql -u root -p epics < epics.sql | ||
(last command took about 19 minutes). | |||
TODO: | |||
understand previous errors | |||
when script 'get_latest_cont.pl' was started it produced the number of errors into log file | |||
'/home/epics/history/log/get_latest_cont.log'; currently script is writing to /dev/null | |||
so log file can be used to see what was wrong |
Revision as of 12:04, 11 January 2007
Resides in CVSROOT=/home/epics/cvs, normally checked out to /home/epics/history and running on clon01 as user epics.
Main script is:
get_latest_cont.pl
It reads EPICS channels and updates epics database on clondb1.
Cron job is checking if main script is running (if not it will try to restart it):
0,15,30,45 * * * * /home/epics/history/www/check_get.pl
CURRENTLY GOES TO /dev/null !!! If started by cron job, log file /home/epics/history/log/get_latest_cont.log will be produced.
Database epics on clondb1 was creates as follwoing:
mysql> create database epics; mysql> grant ALL on epics.* to marki@clon01.jlab.org identified by '********'; mysql> grant ALL on epics2.* to marki@clon01.jlab.org identified by '********'; mysql> grant SELECT,INSERT,UPDATE on epics.* to epics@clon01.jlab.org; mysql> grant SELECT,INSERT,UPDATE on epics2.* to epics@clon01.jlab.org; mysql> grant SELECT on epics.* to clasweb@claspc28.jlab.org; mysql> grant SELECT on epics2.* to clasweb@claspc28.jlab.org;
Web software runs on clasweb machine. It can be found in /u/group/clas/www/clasweb/htbin/epics_history' directory, in particular file 'db_parms.pm' contains the name of database machine (in our case clondb1) and other related attributes.
IMPORTANT: DUMPING FROM clasweb-bck:
cd /home mysqldump -uroot -pe1tocome epics | gzip > /misc/mysql/clasweb-bck/epics.sql.gz mysqldump: Can't get CREATE TABLE for table `hallb_sf_xy560_0_12_1min` (Can't open file: 'hallb_sf_xy560_0_12_1min.MYI'. (errno: 145))
Following command produce error message but runs (took about 33 minutes):
mysqldump -f -l --opt -uroot -pe1tocome epics | gzip > /home/epics.sql.gz mysqldump: Got error: 1016: Can't open file: 'hallb_sf_xy560_0_12_1min.MYI'. (errno: 145) when using LOCK TABLES mysqldump: Can't get CREATE TABLE for table `hallb_sf_xy560_0_12_1min` (Can't open file: 'hallb_sf_xy560_0_12_1min.MYI'. (errno: 145)) mysqldump: Can't get CREATE TABLE for table `hallc_MeV_1min` (Can't open file: 'hallc_MeV_1min.MYI'. (errno: 145)) mysqldump: Can't get CREATE TABLE for table `ioctorus1colonloadPercent_1min` (Can't open file: 'ioctorus1colonloadPercent_1min.MYI'. (errno: 145))
Import into clondb1: type following on clondb1:
cd /database scp clasweb-bck:/home/epics.sql.gz . gunzip epics.sql.gz mysql -u root -p epics < epics.sql
(last command took about 19 minutes).
TODO:
understand previous errors
when script 'get_latest_cont.pl' was started it produced the number of errors into log file '/home/epics/history/log/get_latest_cont.log'; currently script is writing to /dev/null so log file can be used to see what was wrong