- 1. Re: [PATCHv7 0/6] ipvs: Use kthreads for stats (score: 1)
- Author: Julian Anastasov <ja@xxxxxx>
- Date: Sat, 10 Dec 2022 02:47:45 +0200 (EET)
- Hello, There are probably more reasons why kthreads look better: - with kthreads we run code that is read-mostly, no write/lock operations to process the estimators in 2-second intervals - work items
- /html/lvs-devel/2022-12/msg00010.html (12,298 bytes)
- 2. Re: [PATCHv7 0/6] ipvs: Use kthreads for stats (score: 1)
- Author: Pablo Neira Ayuso <pablo@xxxxxxxxxxxxx>
- Date: Fri, 9 Dec 2022 21:58:36 +0100
- OK, then my understanding is that you have requirements to have more control on the kthreads than what the workqueue interface provides. I can see there is WQ_HIGHPRI and WQ_CPU_INTENSIVE flags to si
- /html/lvs-devel/2022-12/msg00009.html (12,026 bytes)
- 3. Re: [PATCHv7 0/6] ipvs: Use kthreads for stats (score: 1)
- Author: Julian Anastasov <ja@xxxxxx>
- Date: Thu, 8 Dec 2022 19:03:44 +0200 (EET)
- Hello, The estimation can take long time when using multiple IPVS rules (eg. millions estimator structures) and especially when box has multiple CPUs due to the for_each_possible_cpu usage that expec
- /html/lvs-devel/2022-12/msg00008.html (10,738 bytes)
- 4. Re: [PATCHv7 0/6] ipvs: Use kthreads for stats (score: 1)
- Author: Pablo Neira Ayuso <pablo@xxxxxxxxxxxxx>
- Date: Thu, 8 Dec 2022 13:17:20 +0100
- Oh wait. I have to hold this back, I have a fundamental question: [PATCHv7 4/6] ipvs: use kthreads for stats estimation uses kthreads, these days the preferred interface for this is the generic workq
- /html/lvs-devel/2022-12/msg00007.html (9,603 bytes)
- 5. Re: [PATCHv7 0/6] ipvs: Use kthreads for stats (score: 1)
- Author: Pablo Neira Ayuso <pablo@xxxxxxxxxxxxx>
- Date: Thu, 8 Dec 2022 13:06:11 +0100
- Series applied to nf-next, thanks.
- /html/lvs-devel/2022-12/msg00006.html (9,407 bytes)
- 6. [PATCHv7 0/6] ipvs: Use kthreads for stats (score: 1)
- Author: Julian Anastasov <ja@xxxxxx>
- Date: Tue, 22 Nov 2022 18:45:58 +0200
- Hello, This patchset implements stats estimation in kthread context. It replaces the code that runs on single CPU in timer context every 2 seconds and causing latency splats as shown in reports [1],
- /html/lvs-devel/2022-11/msg00058.html (17,623 bytes)
This search system is powered by
Namazu