Search String: Display: Description: Sort:

Results:

References: [ +subject:/^(?:^\s*(re|sv|fwd|fw)[\[\]\d]*[:>-]+\s*)*\[PATCH\s+net\-next\]\s+net\:\s+ipvs\:\s+random\s+start\s+for\s+RR\s+scheduler\s*$/: 6 ]

Total 6 documents matching your query.

1. Re: [PATCH net-next] net: ipvs: random start for RR scheduler (score: 1)
Author: Julian Anastasov <ja@xxxxxx>
Date: Fri, 13 May 2022 01:23:52 +0300 (EEST)
Hello, Looks like I forgot to add this example how to reduce code under lock because add/del dest can run in parallel with scheduling, something that is a drawback in this solution: +static void ip_v
/html/lvs-devel/2022-05/msg00010.html (10,671 bytes)

2. Re: [PATCH net-next] net: ipvs: random start for RR scheduler (score: 1)
Author: Julian Anastasov <ja@xxxxxx>
Date: Tue, 10 May 2022 23:18:40 +0300 (EEST)
Hello, Me too, so I created a test for the eyes :) I hope it correctly implements the algorithm. include <stdio.h> include <stdlib.h> include <string.h> /* MAX_STEP: * 4: last has 3 times more chance
/html/lvs-devel/2022-05/msg00006.html (13,100 bytes)

3. Re: [PATCH net-next] net: ipvs: random start for RR scheduler (score: 1)
Author: Menglong Dong <menglong8.dong@xxxxxxxxx>
Date: Tue, 10 May 2022 11:21:17 +0800
Sorry to have missed your message here. Yeah, this is what I mean. And in my case, the directors are local the to client, and each client only have 2 connections. If the 3th connection happens, it wi
/html/lvs-devel/2022-05/msg00004.html (15,689 bytes)

4. Re: [PATCH net-next] net: ipvs: random start for RR scheduler (score: 1)
Author: Menglong Dong <menglong8.dong@xxxxxxxxx>
Date: Tue, 10 May 2022 10:55:59 +0800
Nice description :/ Yeah, prandom_u32_max is a good choice, I'll use it instead. I'm a little afraid that the 'steps' may make the starting dest not absolutely random, in terms of probability. For ex
/html/lvs-devel/2022-05/msg00003.html (14,094 bytes)

5. Re: [PATCH net-next] net: ipvs: random start for RR scheduler (score: 1)
Author: Julian Anastasov <ja@xxxxxx>
Date: Mon, 9 May 2022 21:17:25 +0300 (EEST)
Hello, ...order of added destinations,... ..."long connections are used". Is this a case where small number of connections are used? And the two connections relatively overload the real servers? You
/html/lvs-devel/2022-05/msg00002.html (13,783 bytes)

6. [PATCH net-next] net: ipvs: random start for RR scheduler (score: 1)
Author: menglong8.dong@xxxxxxxxx
Date: Mon, 9 May 2022 20:22:13 +0800
For now, the start of the RR scheduler is in the order of dest service added, it will result in imbalance if the load balance is done in client side and long connect is used. For example, we have cli
/html/lvs-devel/2022-05/msg00001.html (11,665 bytes)


This search system is powered by Namazu