[Pgpool-general] INSERT slow on one system not on another

[Derek Jones]

Dear Marcelo,

Using:

pgpool-II version 2.1 (inamiboshi)


Dear Diego,

 > As a not-that-unrelated note: I understand that the current behaviour
 > does not make much sense, so trying seemingly unrelated changes may
 > lead you to a solution.


Hey ! :-D I like this! Buy ice cream - visit Hawaii - totally unrelated 
- may lead me to a solution ;-) ;-)


Thanks for the suggestion - I know what you mean I've been thinking 
about that - changing the mount options for the drive, looking at kernel 
shared memory, buffering, network connection diffs, etc. (not *totally* 
unrelated I guess).

Although I can't easily change the devel cluster either, what I *might* 
be able to do is clone into a virtual cluster setup with vmware that 
uses the same kernel etc. Of course, then I introduce virtualization but 
at least I can use the same kernel, PXE, etc. (I had this as a 
mini-setup before anyway before committing to physical hardware for a 
development cluster). Now, where did I put that disk space.....

In addition, I realized I had a node down that's on the production 
system for testing, but it might allow me a base to attempt a kernel 
rebuild with a newer kernel and see if I can at least achieve that 
cleanly and keep existing params and modules.

But even then, part of the issue is that I'm teleworking on the 
production cluster, so, any changes that require physical intervention 
(like a reboot gone bad -  would be problematic - would have to travel 
there or get an onsite person to assist. I have a blade management 
interface I can reboot it from remotely normally, but that won't help if 
I get it stuck.)


Kind regards

Derek.