| 1 |
8/01/2013 dcs1x:/home/prodtest/cvs/JSOC/doc/convert_dcs1_to_dcs1x.txt |
| 2 |
|
| 3 |
PRELIMINARY NOTES |
| 4 |
------------------- |
| 5 |
|
| 6 |
We must keep the name dcs1x. |
| 7 |
At the end dcs1 will be retired. |
| 8 |
|
| 9 |
Alert DDS that we are beginning the transition process. |
| 10 |
There may be a time when they cannot transfer data. |
| 11 |
(See if we can mitigate this) |
| 12 |
Any missed data will automatically be re-requested once the |
| 13 |
dcs1x is operational. |
| 14 |
|
| 15 |
Stop all processing on dcs1: |
| 16 |
(First figure out what to do about the warmstandby on dcs2?) |
| 17 |
*Hit Exit on the socdc gui |
| 18 |
*sum_stop_dc |
| 19 |
after do backup below: |
| 20 |
*pg_ctl -D /var/lib/pgsql/dcs1_data stop [-m fast] |
| 21 |
|
| 22 |
Backup the dcs1 db. Remember there is no drms on dcs1. |
| 23 |
Make sure all the current rows in dcs1x db have been purged. |
| 24 |
|
| 25 |
We could us this: |
| 26 |
Begin the backup process on dcs1: |
| 27 |
> psql -p 5431 -c "select pg_start_backup('2013-07-19')" |
| 28 |
Create a tar archive of the system data directory: |
| 29 |
> tar -czvf dcs1_backup.tar.gz dcs1_data/ |
| 30 |
Stop the backup process on dcs1: |
| 31 |
> psql -p 5431 -c "select pg_stop_backup()" |
| 32 |
|
| 33 |
or since there are only 9 sums tables to transfer, it might |
| 34 |
be better just to copy like so: |
| 35 |
jsocdc=# copy sum_file to '/tmp/sum_file.copy'; |
| 36 |
and on dcs1x after scp the file over: |
| 37 |
hmidb_sums=# copy sum_file from '/tmp/sum_file.copy'; |
| 38 |
The files to copy from dcs1 are: |
| 39 |
Schema | Name | Type | Owner |
| 40 |
--------+------------------+----------+---------- |
| 41 |
public | sum_ds_index_seq | sequence | postgres |
| 42 |
public | sum_file | table | postgres |
| 43 |
public | sum_group | table | postgres |
| 44 |
public | sum_main | table | postgres |
| 45 |
public | sum_open | table | postgres |
| 46 |
public | sum_partn_alloc | table | postgres |
| 47 |
public | sum_partn_avail | table | postgres |
| 48 |
public | sum_seq | sequence | postgres |
| 49 |
public | sum_tape | table | postgres |
| 50 |
(9 rows) |
| 51 |
|
| 52 |
|
| 53 |
The drms tables are in place on dcs1x. The defined series are: |
| 54 |
Name Archive Retention UnitSize |
| 55 |
-------------------- ------- --------- -------- |
| 56 |
hmi.lev0_isp_0021 0 60 1 |
| 57 |
hmi.lev0_seq_0011 0 60 1 |
| 58 |
hmi.tlm 1 5 1 |
| 59 |
hmi.lev0a -1 60 12 |
| 60 |
|
| 61 |
TBD: Notes on moving the network connection and IP addr. And how |
| 62 |
we stop and start the dataflow. |
| 63 |
|
| 64 |
Copy the dcs1:/SUM1 |
| 65 |
drwxrwsr-x 61149 root SOI 1609728 Jul 19 10:43 SUM1/ |
| 66 |
to dcs1x:/SUM1 |
| 67 |
drwxrwsr-x 51996 sumsadmin sums 1175552 Jun 27 01:11 SUM1/ |
| 68 |
|
| 69 |
|
| 70 |
Populate the T50 with the original dcs1 tapes. |
| 71 |
|
| 72 |
NFS /data on dcs1x to cl1n001 |
| 73 |
|
| 74 |
To start dcs1x, see: |
| 75 |
/home/prodtest/cvs/JSOC/doc/new_dcsx_run.txt |
| 76 |
|
| 77 |
Enter the crontab -e with the jobs from dcs1. |
| 78 |
|
| 79 |
NOTE: This doc s/b developed into a checklist to be followed. |
| 80 |
|
| 81 |
NOTE: The dcs2x has both a hmidb[_sums] and a aiadb[_sums], so it |
| 82 |
can temporarily take on either role if dcs0x or dcs1x is out of service. |
| 83 |
The concept is that data would be reprocessed when the main machine is back, |
| 84 |
and tape archiving would proceed from where it left off. |
| 85 |
We should look closer at this, but an attempt has been made to simplify this |
| 86 |
host backup from what we had previously (conflicting db names and port |
| 87 |
numbers and /SUMs names, etc. have been eliminated). |
