Return-Path: X-Original-To: archive-asf-public-internal@cust-asf2.ponee.io Delivered-To: archive-asf-public-internal@cust-asf2.ponee.io Received: from cust-asf.ponee.io (cust-asf.ponee.io [163.172.22.183]) by cust-asf2.ponee.io (Postfix) with ESMTP id 2D73A200D24 for ; Tue, 10 Oct 2017 04:05:55 +0200 (CEST) Received: by cust-asf.ponee.io (Postfix) id 2BE161609E0; Tue, 10 Oct 2017 02:05:55 +0000 (UTC) Delivered-To: archive-asf-public@cust-asf.ponee.io Received: from mail.apache.org (hermes.apache.org [140.211.11.3]) by cust-asf.ponee.io (Postfix) with SMTP id F07951609CE for ; Tue, 10 Oct 2017 04:05:53 +0200 (CEST) Received: (qmail 9975 invoked by uid 500); 10 Oct 2017 02:05:52 -0000 Mailing-List: contact dev-help@airavata.apache.org; run by ezmlm Precedence: bulk List-Help: List-Unsubscribe: List-Post: List-Id: Reply-To: dev@airavata.apache.org Delivered-To: mailing list dev@airavata.apache.org Received: (qmail 9965 invoked by uid 99); 10 Oct 2017 02:05:52 -0000 Received: from pnap-us-west-generic-nat.apache.org (HELO spamd2-us-west.apache.org) (209.188.14.142) by apache.org (qpsmtpd/0.29) with ESMTP; Tue, 10 Oct 2017 02:05:52 +0000 Received: from localhost (localhost [127.0.0.1]) by spamd2-us-west.apache.org (ASF Mail Server at spamd2-us-west.apache.org) with ESMTP id 99BD51A50EE for ; Tue, 10 Oct 2017 02:05:51 +0000 (UTC) X-Virus-Scanned: Debian amavisd-new at spamd2-us-west.apache.org X-Spam-Flag: NO X-Spam-Score: 3.579 X-Spam-Level: *** X-Spam-Status: No, score=3.579 tagged_above=-999 required=6.31 tests=[DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, HTML_MESSAGE=2, KAM_LINEPADDING=1.2, RCVD_IN_DNSWL_NONE=-0.0001, RCVD_IN_MSPIKE_H3=-0.01, RCVD_IN_MSPIKE_WL=-0.01, RCVD_IN_SORBS_SPAM=0.5, SPF_PASS=-0.001] autolearn=disabled Authentication-Results: spamd2-us-west.apache.org (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com Received: from mx1-lw-us.apache.org ([10.40.0.8]) by localhost (spamd2-us-west.apache.org [10.40.0.9]) (amavisd-new, port 10024) with ESMTP id GcDTF4RNn2FF for ; Tue, 10 Oct 2017 02:05:48 +0000 (UTC) Received: from mail-oi0-f50.google.com (mail-oi0-f50.google.com [209.85.218.50]) by mx1-lw-us.apache.org (ASF Mail Server at mx1-lw-us.apache.org) with ESMTPS id 9A29B5FD1B for ; Tue, 10 Oct 2017 02:05:48 +0000 (UTC) Received: by mail-oi0-f50.google.com with SMTP id h200so15794197oib.4 for ; Mon, 09 Oct 2017 19:05:48 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=mime-version:in-reply-to:references:from:date:message-id:subject:to; bh=Wvdu1wnZRTtVE/mp3+CyLs2cbUjSTIVi/FkRLYt/7Uk=; b=LKlA+9uYjFL67AUV7qn4XYywfmwrfy3NWKKyzLpQbxgim4XPXQuQm5vjyG7wPoEkur kZxhJ4tiTjvSDlPBF4iwyyuzqfhFspa36p0vC567EktYY6eXGSy2FDIdC7pfzpBhh7tD mSr8xOKoKOhx6hAPrntNdV3qcUp95X+bo5qePdOXHlbE/RkRICEAMGAHvdkEbO4xZ4wK /n4QtEC3nLr2BWKsOh2caoC1I6/yMCWwNdL9/4DEyBD5mk9QsT4RLL6v2kF/tKgYbazU N7zRUKhMQ2Zl2RSZPvV6i9CFQMSjYvdcO878ZhRzfy9nPeAoR96A84I8Y8n+SBQ9TFlW LHkg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to; bh=Wvdu1wnZRTtVE/mp3+CyLs2cbUjSTIVi/FkRLYt/7Uk=; b=bqotV2xLsiC76B+TwhwxdD4pGY7eOGB3pWfwvqDKY6sTdlC/KX2WV2mCirIBSwNTCD EzXFKpKQSqzmv+O3V+8gVPb1C7QcqnqUyEHwg4/XJmnQ3efB2r++3gK8NrfG0pkVI4Zz YWZCmh0xb9BWL8Xipjcw2w1cudPoYoVobvuJ9C1zvuEz1IAaKxJ4O8gM4OV82HvinXO3 z5KMkuhyAdcKQHUpEjI7h8ESeZzfbVCAevEvby+qt1391PzGE6lcbdui1tH57jbsPXVD Ir89cqWsZEyVylpxo3t1hOHd+mYTISOBtJvmPakEh2jCYQRrrrsxhTxRi0tcQ7pr2yRG N2+w== X-Gm-Message-State: AMCzsaWuttCSGzcEsF7wPwUNlIDU8nLkCbmNsJb+SSCYObbQsBwXCLWx evoIxAwMdJI+OlTYHbBrWMXbB2ZsL8eioOu2wMg= X-Google-Smtp-Source: AOwi7QAM21kNEN0Eu3pDAxiPGWV4VdLYcCzQufUtHQgh+/TnZA6a2sveRR75u1f4omdppLEBUBPn8Cte1ciJx+ei0XE= X-Received: by 10.157.40.101 with SMTP id h34mr722748otd.236.1507601147739; Mon, 09 Oct 2017 19:05:47 -0700 (PDT) MIME-Version: 1.0 Received: by 10.157.55.244 with HTTP; Mon, 9 Oct 2017 19:05:27 -0700 (PDT) In-Reply-To: References: <7EE45E05-666E-47A2-9735-60E8D11081CA@iu.edu> From: Jerrin Suresh Date: Mon, 9 Oct 2017 22:05:27 -0400 Message-ID: Subject: Re: Working on Load Balancers To: dev@airavata.apache.org Content-Type: multipart/alternative; boundary="001a113df514dfc463055b27bea1" archived-at: Tue, 10 Oct 2017 02:05:55 -0000 --001a113df514dfc463055b27bea1 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Thanks Gourav for the update. So, we will have to implement multiple load balancers in-order to avoid a single point of failure. However, now I am facing an issue with binding the ports in consul and ha-proxy. Within the openstack, there is no provision to access the public ip of an instance. Hence, ha-proxy, consul-template and consul are accessing the private IP of the portal server. I haven't moved the setup to production. This was the case when I tried implementing the dummy concept on TACC openstack. So, is there a way in which I can access the floating IP of instances within an instance? Thanks, Jerrin On Tue, Oct 3, 2017 at 3:39 PM, Shenoy, Gourav Ganesh wrote: > Hi Jerrin, > > > > Avoiding a single point of failure for the load balancer is tricky. But > typically, a good approach is to have multiple load balancers (for HA) an= d > a floating IP which maps to an =E2=80=9Cactive=E2=80=9D load balancer =E2= =80=93 this is assuming we > have an Active-Passive type HA cluster for the load balancer. Container > orchestration engines such as Kubernetes/DCOS provide such features > out-of-the-box; same goes with public clouds such as AWS or GCE. Privatel= y > managed cloud like OpenStack will need some complex configuring to enable > such DNS mapping. > > > Hope this helps! > > > > Thanks and Regards, > > Gourav Shenoy > > > > *From: *Jerrin Suresh > *Reply-To: *"dev@airavata.apache.org" > *Date: *Tuesday, October 3, 2017 at 3:31 PM > *To: *"dev@airavata.apache.org" > *Subject: *Re: Working on Load Balancers > > > > The main idea of implementing a load balancer is to balance the load to > the portal and remove any single point of failure. However, after > implementing the load balancer, the load balancer itself becomes a single > point of failure. > > > > Any thoughts regarding the same? Should we have multiple load balancers > involved, i.e a master and a slave or a dual master setup? > > > > > > > > On Mon, Oct 2, 2017 at 10:51 AM, Christie, Marcus Aaron > wrote: > > No plans. But if it makes sense to do so we could. > > > > On Oct 2, 2017, at 10:03 AM, Jerrin Suresh wrote= : > > > > Just a quick question, do you have plans of dockerizing the Django portal= ? > > > > > > ~jerrin > > > > On Fri, Sep 29, 2017 at 10:58 PM, Jerrin Suresh > wrote: > > Hi Marcus, > > > > Thanks for the update. I shall set the load balancer up for a sample > application, and am planning to use haproxy for the project. > > > > Regards, > > Jerrin > > > > On Fri, Sep 29, 2017 at 9:11 AM, Christie, Marcus Aaron > wrote: > > Jerrin, > > > > What specifically do you want to work on regarding load balancing? > Depending on what you want to work on, load balancing a web application i= s > fairly independent of the web application framework. So I would say for > now you could just try load balancing any web application, even a simple > little one written in either Laravel or Django (or another framework). > > > > We run the PHP PGA and Django PGA in Apache HTTPD server. So it would be > good if the load balancing work you do works with Apache HTTPD. However, > I=E2=80=99m open to using something else like nginx or whatever if you ca= n make a > good case for it. > > > > Thanks, > > > > Marcus > > > > On Sep 28, 2017, at 11:25 AM, Jerrin Suresh > wrote: > > > > Hi, > > > > I am working on setting up Load Balancers for Apache Airavata. However, a= s > the PGA is being moved to the Django framework is it better to work on th= e > PGA branch or should I work on the Django branch? > > > > > ~jerrin > > -- > > MS CS Fall-2018 > > Indiana University > www.linkedin.com/in/jerrinsuresh > > > > > > > > > > -- > > MS CS Fall-2018 > > Indiana University > www.linkedin.com/in/jerrinsuresh > > > > > > > > -- > > MS CS Fall-2018 > > Indiana University > www.linkedin.com/in/jerrinsuresh > > > > > > > > > > -- > > MS CS Fall-2018 > > Indiana University > www.linkedin.com/in/jerrinsuresh > > > --=20 MS CS Fall-2018 Indiana University www.linkedin.com/in/jerrinsuresh --001a113df514dfc463055b27bea1 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable
Thanks Gourav=C2=A0for the update. So, we will have to imp= lement=C2=A0multiple load=C2=A0balancers in-order to avoid a single point o= f failure.=C2=A0

However, now I am facing an issue with = binding the ports in consul and ha-proxy. Within the openstack, there is no= provision to access the public ip=C2=A0of an instance. Hence, ha-proxy, co= nsul-template and consul are accessing the private IP of the portal server.= =C2=A0

I haven't moved the setup to production= . This was the case when I tried implementing the dummy concept on TACC ope= nstack. So, is there a way in which I can access the floating IP of instanc= es within an instance?

Thanks,
Jerrin

On Tue, = Oct 3, 2017 at 3:39 PM, Shenoy, Gourav Ganesh <goshenoy@indiana.edu= > wrote:

Hi Jerrin,

=C2=A0

Avoiding a single point of failure for the load balancer is tricky. But ty= pically, a good approach is to have multiple load balancers (for HA) and a = floating IP which maps to an =E2=80=9Cactive=E2=80=9D load balancer =E2=80=93 this is assuming we have an Active-Passive type HA= cluster for the load balancer. Container orchestration engines such as Kub= ernetes/DCOS provide such features out-of-the-box; same goes with public cl= ouds such as AWS or GCE. Privately managed cloud like OpenStack will need some complex configuring to enable such DNS= mapping.


Hope this helps!

=C2=A0

Thanks and Regards,

Gourav Shenoy

=C2=A0

F= rom: Jerrin Suresh <jerrinsuresh@gmail.c= om>
Reply-To: "dev@airavata.apache.org" <dev@airavata.apache.org>
Date: Tuesday, October 3, 2017 at 3:31 PM
To: "dev@airavata.apache.org" <dev@airavata.apache.org>
Subject: Re: Working on Load Balancers

=C2=A0

The main idea of implementing a load balancer is to = balance the load to the portal and remove any single point of failure. Howe= ver, after implementing the load balancer, the load balancer itself becomes= a single point of failure.

=C2=A0

Any thoughts regarding the same? Should we have mult= iple load balancers involved, i.e a master and a slave or a dual master set= up?

=C2=A0

=C2=A0

=C2=A0

On Mon, Oct 2, 2017 at 10:51 AM, Christie, Marcus Aa= ron <machrist@iu.ed= u> wrote:

No plans. But if it makes sense to do so we could. <= u>

=C2=A0

On Oct 2, 2017, at 10:03 AM, Jerrin Suresh <jerrinsuresh@gmail.c= om> wrote:

=C2=A0

Just a quick question, do you have plans of dockeriz= ing the Django portal?

=C2=A0

=C2=A0

~jerrin

=C2=A0

On Fri, Sep 29, 2017 at 10:58 PM, Jerrin Suresh <= jerrinsuresh@gm= ail.com> wrote:

Hi Marcus,

=C2=A0

Thanks for the update. I shall set the load balancer= up for a sample application, and am planning to use haproxy=C2=A0for the p= roject.

=C2=A0

Regards,

Jerrin

=C2=A0

On Fri, Sep 29, 2017 at 9:11 AM, Christie, Marcus Aa= ron <machrist@iu.ed= u> wrote:

Jerrin,

=C2=A0

What specifically do you want to work on regarding l= oad balancing?=C2=A0 Depending on what you want to work on, load balancing = a web application is fairly independent of the web application framework.= =C2=A0 So I would say for now you could just try load balancing any web application, even a simple little one written in ei= ther Laravel or Django (or another framework).

=C2=A0

We run the PHP PGA and Django PGA in Apache HTTPD se= rver.=C2=A0 So it would be good if the load balancing work you do works wit= h Apache HTTPD. However, I=E2=80=99m open to using something else like ngin= x or whatever if you can make a good case for it.

=C2=A0

Thanks,

=C2=A0

Marcus

=C2=A0

On Sep 28, 2017, at 11:25 AM, Jerrin Suresh <jerrinsuresh@gmail.= com> wrote:

=C2=A0

Hi,

=C2=A0

I am working=C2=A0on setting up Load Balancers for A= pache Airavata. However, as the PGA is being moved to the Django framework = is it better to work on the PGA branch or should I work on the Django branc= h?

=C2=A0


~jerrin

--

MS CS Fall-2018

Indiana University
www.linkedin.com/= in/jerrinsuresh

=C2=A0

=C2=A0



=C2=A0

--

MS CS Fall-2018

Indiana University
www.linkedin.com/= in/jerrinsuresh

=C2=A0



=C2=A0

--

MS CS Fall-2018

Indiana University
www.linkedin.com/= in/jerrinsuresh

=C2=A0

=C2=A0



=C2=A0

--

MS CS Fall-2018

Indiana University
www.linkedin.com/= in/jerrinsuresh

=C2=A0




--
MS CS Fall-2018
Indiana University
www.linkedin.com/in/jerrinsuresh

--001a113df514dfc463055b27bea1--