Re: Table freeze when size exceeds innodb buffer pool size

[newbie Shai]

Hi Peter,

1. For this is any minimum benchmark or allowed scans per seconds ?Select_scan / Uptime_since_flush_status --> table scans/sec. What this means is that every second there is a table scan is it?

2. I am a bit confuse withe select_scan and Handler_read_rnd_next ? Arent both related ?

3. The problem with the upgrade I got get the permission from the user and on top of this is a live system which is supported for gps so data keeping coming on live.

4. I dont get you on this innodb_os_log_written / Uptime_since_flush_status / innodb_log_files_in_group, rounded uo to a 32MB boundary. What's best is often about 30 mins of writes. ? Is it 24123085824 /898882/2 I get the answer is some thing like 13418 what does this states?

5. If I leave the innodb_buffer_pool will it have any impact? Actually how it correspond to a particular table or to the whole database to be able to accomodate into the memory?

Besides this variable what else normally experts will do to hunt for the clues?