February 01, 2005
"I suppose that you'll be needing it now."

I won't go into much detail yet over yesterdays thirteen hour epic, but it was a narrowly avoided catastrophe.

Suffice it to say that we needed a copy of Lumigent Log Explorer, sharpish. I gave Solution Data a bell at around five. "Can you give me some idea as to the lead time on a copy of Log Explorer, please?"

"I suppose that you'll be needing it now, won't you? You can download it now, and I can get you a key in, oh, an hour or so." Clearly, they know their market!

"Err, yes."

They were as good as their word. So, a big thank you to Andrew Hulley over at Solution Data.

As for Lumigent Log Explorer, well, it works. ;-) It does what it says on the tin, and the interface was good enough that we were able to do what we needed to do, despite the fact that our first exposure to the tool was when we were using it to fix a critical problem. Not pretty, and a bit complex, but then it's not exactly an end user tool. You don't need Log Explorer, but if you do need it, you need it bad.

One other thing. If your backup strategy depends on something as complex and delicate as SQL Server Replication, then it's not a robust backup strategy. ;-)

Posted to Software development by Simon Brunning at February 01, 2005 10:40 AM

Well done guys.

And SQL Server Replication is VERY robust... if you set it up right ;-)
This is a genuine example of your non-MS bias here mate, tut tut, I'm actually dissapointed in you over this throw away statement, because it's alkso technically untrue.
For example, if you set up N servers doing complex rule driven ywo-way replication between them, throw lots of data at it (megs or 100's of megs), and pull the plug out of one or some or all machines randomly during a massive re-synch, the data will STILL be good, and the transaction will complete ok, when you power the machines back up.
In my book, that's robust.
If you've turned the replication on and configured it rigth, which is easy.

Posted by: Mark Matthews on February 2, 2005 09:38 AM

The fact is that replication stopped, we didn't know about it, and we nearly lost data because of it. There's no point telling me that replication doesn't go wrong - it did, and Tulna, El P and I were at work 'till 10:30 on Monday night sorting it all out. That's not bias, that's a fact.

And as for setting it up correctly - it was you that set it up!

Your backup strategy should be very robust - and simple enough that you *know* it's right. Our's wasn't. Lesson learned.

Posted by: Simon Brunning on February 2, 2005 11:25 AM
Post a comment

Email Address:



Remember info?