[Glass] Transaction Logs so large ....

[Dale Henrichs via Glass]

On 10/24/2016 04:22 AM, Paul Baumann via Glass wrote:
>
> Tranlog files at first have transaction sequence not yet represented 
> in the random-access database extent files. Sequential files are 
> faster to update and changes can be replayed to extents of different 
> ages (like from restore or cloning). Tranlog files that only contain 
> transactions older than the last Checkpoint Commit can be archived. A 
> checkpoint commit is the point where extents now contain the tranlog 
> changes. You do not need to keep full tranlog files older than the 
> checkpoint commit record in the extents backup that you may wish 
> restore. GS DBAs write shell scripts to manage these things, knowing 
> what tranlog files are no longer useful to keep.
>
> If you already know this stuff then your question would be related to 
> checkpoint commits not happening in a timely manner. How that can 
> happen and how to improve that is a much deeper discussion, but is 
> usually remedied by changing gems to not stay with old views of the 
> database. That can mean more frequent aborts for example, and starting 
> a transaction a short time before making changes to commit. Just one 
> gem staying in-transaction for a long time can hold up checkpoint 
> commits. A developer logged into (and in-transaction) a very active 
> database can hold up checkpoints causing a backlog of transactions not 
> yet applied to extent files.
>
Paul, a CR backlog dos not cause tranlogs to grow ... CR backlog shows 
up as extent growth not tranlog file growth...

Dale