Finally blog (/ blogs / 2013/06/28 / efficiency queue, but where /) we discussed the old and the new queue mechanism in NetScaler and now let us understand how the new implementation works.
control queues to the application or the layer of vserver should strictly use in waiting and logic-tail so that it controls the unified distribution across file services and also provides protection functions work as expected.
en-tail Trigger
queue up ... sounds simple, but it's there, we need to ensure that the Line waiting is effective and performance oriented beyond the management functionality ... let us look at the use cases:
1) the resources available on the service
- Let LB get treatment and conn is given as normal service
2) the resources are not available on the standby service and file EXIST on vserver
- Allocate memory for the queue in vserver
- In- queue at the vserver
3) Resources are not available on service, but EXIST queue on vserver
- -in queue at the vserver
4) ARE tail on vserver but resources available
- in-queue the current connection
- event queue Trigger on the head of the queue
de-tail Trigger
Clearing the queue is even easier to the extent that the capacity exists to end services. Here fast flux:
1) Login New customer
- In this connection tail
- De-existing customer queue ahead of the queue
2) the server triggered queue without Persistence
- If the server side of the connection is free
- Surge account on the server is 0
- Put on the same server connection client that triggered queue
3) server triggered waiting and persistence file
- if the fault we have always-tail corresponding customer
- is no corresponding client is waiting and then we take the highest priority customer
4) lead Timer waiting on the basis
- All one second, we'll wait to resume LB-file and a client connection
in short this new technology queue will help end users and owners of large-scale applications ... more reasons to visit NetScaler 10.1 release 
0 Komentar