[Date Prev][Date Next]
[Chronological]
[Thread]
[Top]
Antw: Re: Number of requests served in ideal condition by version 2.4.46 with lmdb backend.
- To: "Saurabh Lahoti" <saurabh.astronomy@gmail.com>, <openldap-technical@openldap.org>,"Howard Chu" <hyc@symas.com>
- Subject: Antw: Re: Number of requests served in ideal condition by version 2.4.46 with lmdb backend.
- From: "Ulrich Windl" <Ulrich.Windl@rz.uni-regensburg.de>
- Date: Wed, 01 Aug 2018 07:49:10 +0200
- Content-disposition: inline
- In-reply-to: <025307a1-4de1-3299-e935-83ba08c11586@symas.com>
- References: <CAB5-MqoiF+_2hteqW1Twv8-Wij6f9EaFTA7AcN7_OKdvGgCWog@mail.gmail.com> <025307a1-4de1-3299-e935-83ba08c11586@symas.com>
>>> Howard Chu <hyc@symas.com> schrieb am 31.07.2018 um 21:24 in Nachricht
<025307a1-4de1-3299-e935-83ba08c11586@symas.com>:
> Saurabh Lahoti wrote:
>> Dear All,
>>
>> For preparing a technical scope of OpenLDAP; do we have any maximum
> threshold of requests served per second by single instance of 2.4.46 with
> lmdb backend..?
>
> No. The request rate is entirely determined by your network interface speed
> and number of CPUs.
> LMDB search performance scales linearly with CPU cores.
Wouldn't the memory bandwidth be the actual limiting factor (thinking of some 24 core CPU) when running queries via "localhost"?
>
> --
> -- Howard Chu
> CTO, Symas Corp. http://www.symas.com
> Director, Highland Sun http://highlandsun.com/hyc/
> Chief Architect, OpenLDAP http://www.openldap.org/project/