| Author |
Topic |
|
tapaswani86
Starting Member
26 Posts |
 Posted - 2010-11-25 : 03:32:29
|
| Hi all,what should be my backup Strategy for 8tb database?Thanks,Tapaswani |
|
|
nigelrivett
Master Smack Fu Yak Hacker
3385 Posts |
Posted - 2010-11-25 : 04:17:11
|
| depends on how it is being used.I always thinks of a baseline asFor batch systems think of full weekly and diffs daily or full dailyFor oltp the same but with hourly tr logsFor something this size think of partitioning if possible. Put old data on read only partitions so they are static then you only need to back up the current partition.Do you not have anyone that is used to dealing with systems of this size? If not I think you should consider finding someone unless this is static data. You will also probably have problems accessing the data.==========================================Cursors are useful if you don't know sql.SSIS can be used in a similar way.Beer is not cold and it isn't fizzy. |
 |
|
|
tapaswani86
Starting Member
26 Posts |
Posted - 2010-11-25 : 04:32:23
|
| Thanks |
|
|
|
Lumbago
Norsk Yak Master
3271 Posts |
Posted - 2010-11-25 : 05:04:05
|
| To have an 8TB database and ask about backup strategies I'd say you're in over your head.- LumbagoMy blog (yes, I have a blog now! just not that much content yet) -> www.thefirstsql.com |
|
|
|
Kristen
Test
22859 Posts |
Posted - 2010-11-25 : 11:46:18
|
| The folk processing the huge astronomical database have no backup - they figure they can rebuild the database from the original input files in a couple of weeks (of solid processing!!) |
|
|
|
nigelrivett
Master Smack Fu Yak Hacker
3385 Posts |
Posted - 2010-11-25 : 11:58:42
|
| I worked on a system like that.I pointed out that IDs were allocated to the source data which were referenced in all the rest of the systems that used the data - it would have meant rebuilding the data for all of the dependent systems not just reloading the data. This was a few hundre TB of source data.This was after pointing out the errors in the loads of the curent source data. Wasn't too popular when I said that basing a companies data on a system that covered up load errors wasn't a good idea.==========================================Cursors are useful if you don't know sql.SSIS can be used in a similar way.Beer is not cold and it isn't fizzy. |
|
|
|
Kristen
Test
22859 Posts |
Posted - 2010-11-25 : 12:57:28
|
| Being popular is vastly overrated, eh?!! |
|
|
|
tkizer
Almighty SQL Goddess
38200 Posts |
|
|
Lumbago
Norsk Yak Master
3271 Posts |
Posted - 2010-11-26 : 03:01:33
|
A few hundred TB of data...that is quite a bit! I've never even crossed the TB boundary...yet  - LumbagoMy blog (yes, I have a blog now! just not that much content yet) -> www.thefirstsql.com |
|
|
|
nigelrivett
Master Smack Fu Yak Hacker
3385 Posts |
Posted - 2010-11-26 : 04:15:40
|
| Not even if you include runaway tr logs and tempdb?==========================================Cursors are useful if you don't know sql.SSIS can be used in a similar way.Beer is not cold and it isn't fizzy. |
|
|
|
Lumbago
Norsk Yak Master
3271 Posts |
Posted - 2010-11-26 : 05:35:22
|
| Nope...the largest so far is close to 600GB, but that would be active production data for a phone company so it was still somewhat of a challenge. I've been mostly working with small companies though, 200 employees tops...- LumbagoMy blog (yes, I have a blog now! just not that much content yet) -> www.thefirstsql.com |
|
|
|
|
Backup strategy