One Way Sync +2,500,000 Header Db / +7,500,000 Detail Db

[Jeanswhse]

I post this last year and I haven't had a chance to complete this task. Also, I'm waiting for lastest version.

IBM xSeries 236 4G RAM, Windows 2003 Server, running mysql 5.0.18.

IBM xSeries 235 4G RAM, Suse Linux Enterprise 10.1, running mysql 5.1.35 (Backup Server)

Both servers are on a company's LAN.

mysql's dbs have +2,500,000 (sales header db) & +7,500,000 (sales detail db) records now, and has been updated constantly on a Windows 2003 server from remote stores.

Using sqlyog Enterprise 8.12.

I tested it on a same, mirror and small dbs on daily basic and it's OK and no error.

Does anyone have any experience who wants to share on sync a big dbs using 8.12?

Any issue on different mysql 5.0.18 (source) vs 5.1.35 (target) ?

I want to see how long it will take.

I plan to do this sometime, next week. 🙁

Alfred

[peterlaursen]

We cannot rell how much time it wil take to sync. But it is not the number or rows that matters most, but rather what operations (INSERTS, UPDATES, DELETES) will need to be performed (if there is nothing to do it will take a few minutes only). So the first sync may take considerable time, but sync job following after may take much less time. So the general advice would be that sync job should be scheduled so frequently that sync time is acceptable.

[Jeanswhse]

Thank for your reply!

The first sync from source db to empty target db, how long do you think that will be? 1-2 hours?

Alfred

[peterlaursen]

When target table is empty we are just doing a plain copy (there are no data to compare). It will be almost as fast as importing a dump if network is fast. But I cannot give numbers.

[Author]

Posts