Cannot Sync Databases While Ignoring Structures?

[zentode]

I need to perform a one way sync from one servers db to anothers.

The problem is, I am having issues due to the tables having different structures.

The source is an administrative database, the target is a “site side” database.

The tables contain some different columns but, the previous admin was able to sync with navicat for mac but, navicat for pc does not have the ignore structure feat. So, I am trying SQLyog to circumvent buying a mac and navicat for mac!

I need to be able to sync the databases while ignoring structure differences.

Adding rows, syncing the columns fields that match.

Even if I manually select the columns, I still get the error.

Field count not equal for '`sometable`' table.

Can SQLyog allow me to sync these dbs while ignoring structure diffs or am I stuck here?

[peterlaursen]

We do not support syncing tables with different structures – except for

* tables/columns with different charset/collation (but use it at you own risk – only with English/plain ASCII different charsets will work properly!)

* tables with different  indexes (except for the Primary Key that must be identical)

* some rarely used options like MAX_ROWS, MIN_ROWS etc.

SQLyog Data Sync uses checksums calculated on a complete row (mostly).  There is a option, but  it defines on what columns checksum is calculated only. Even if option is used all columns will be synced. 

We do not have any plans to change the implementation in this respect. *sync* means *make identical* and if even structure is not identical *sync* makes no sense to us.

[zentode]

'peterlaursen' wrote:

We do not support syncing tables with different structures – except for

* tables/columns with different charset/collation (but use it at you own risk – only with English/plain ASCII different charsets will work properly!)

* tables with different  indexes (except for the Primary Key that must be identical)

* some rarely used options like MAX_ROWS, MIN_ROWS etc.

SQLyog Data Sync uses checksums calculated on a complete row (mostly).  There is a option, but  it defines on what columns checksum is calculated only. Even if option is used all columns will be synced. 

We do not have any plans to change the implementation in this respect. *sync* means *make identical* and if even structure is not identical *sync* makes no sense to us.

Thanks,

The deal is the admin side of this installation has table columns that are not used or even included on the live sites side.

The way this site was set up is with an “admin” database where the administrators add product listings. When they are done adding, the admin db was “synced” to the live site db. Inserting rows and populating matching columns with the relevant data.

I guess there was some reason for this but, that is in the past. I have this dropped in my lap and there are over 150,000 listings in the db with more coming in daily. THERE IS NO WAY TO “SYNC” IT and they will go out of business if I cannot do this for them. Looks like I gotta buy an apple and use navicat until I can write a new backend for this beast.

Navicat for mac will do this “out of non-sync” but, not with their pc version. I dislike navicat and was hoping your app could work for me.

Oh well, at least I get to play with a mac for a while.

Please keep me informed if you intend to allow a “ignore structure” when syncing checkbox in your software. I will jump right in and purchase.

Thanks for your time.

[peterlaursen]

I am not sure I would dare syncing live data and 'ignore structure'. You may get NULL replaced with ZERO (what would completely break certain applications) and truncation of data is very likely. I realize that there may be scenarios where a 'sufficiently smart' DBA can handle it, but we cannot ignore the risk of data loss if we distribute such for the public – because then it would be us that would go out of business,

I can only repeat that you shall not expect this implemented in SQLyog Data Sync in foreseeable future.

[ken]

I remember asking about this years ago, same answer…now at least I know why.

Thanks.

[Author]

Posts