Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Beowulf Cluster is very slow. Suggestions needed to increase the speed.

25 views
Skip to first unread message

suresh kannan

unread,
Oct 7, 2014, 12:00:02 AM10/7/14
to
I am an Indian student in suwon, korea. I built a Beowulf cluster (system information below) with four systems in our lab for our simulation work with the help of good tutorials. In those tutorials they have mentioned all the system should have static ip addresses. Unfortunately, in all our labs we have been provided with dynamic ip address[5 ips for 15 members in three separate labs]. I have requested four more ip's from our university system admin. Due to the language problems, i conveyed the requirement through my korean lab mate and i dont know the reason why he denied us the static ip. So i found another way to skip this procedure http://www.reddit.com/r/linuxquestions/comments/2gubad/why_static_ip_address_is_necessary_for_linux/.

Someone suggested to use a router (one static ip) and set static ip for the four computers through a router. I did that and it worked. However, the cluster is very slow. For instance If i submit my simulation job in a single computer [4 core processor], it takes 2 months to complete a specific job. Although, if i connect 4 systems it shows it take 6 months to complete the same job. It is actually using 10 core processor [3,3,2,2-100% each]. I used TOP command to see how much processor the head and other nodes are using. I have used openMPI to do parallel the systems. I am using GROMACS (Parallelization based on MPI has been part of this software). I followed a parallel configuration for the Gromacs with the help of this tutorial http://flakrat.blogspot.kr/2013/04/how-to-compile-gromacs-461-with-openmpi.html. After reading few posts http://www.reddit.com/r/linuxquestions/comments/2gbgbg/what_would_be_the_best_linux_distro_for_folding/ i suspected the network router might be an issue.

Can you suggest me how can i troubleshoot this problem? Some one suggested to use 2 network ports and make linux as a router and use a gigabitswitch to get the speed. However, we dont have 2 network ports system. If this is compulsory i can buy network ports (USB one).

Where do i start now?

Can i make my head node as a router and use USB network port (for the second network port) and connect to a gigabitswitch (any model suggestion?) to connect other nodes. I dont know much about networking stuffs. It will be helpful if any experts can suggest to troubleshoot this issue.

Thank you for your time.

regards

Suresh


System Informations

Head node Processor : Intel core i3
RAM : 1 GB
No. of processor : 4
Network cards : 03:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller (rev 06)
System company : Samsung
Architecture : x86_64
OS flavour : Linux Mint 17 Qiana

Node1 Processor : Intel Quad core
RAM : 3 GB
No. of processor : 4
Network cards : 03:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller (rev 02)
System company : TG DREAMSYS
Architecture : x86_64
OS flavour : Linux Mint 17 Qiana

Node2 Processor : Intel core i3
RAM : 1 GB
No. of processor : 4
Network cards : 03:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller (rev 06)
System company : Samsung Architecture : x86_64
OS flavour : Linux Mint 17 Qiana

Node3 Processor : Intel core i3
RAM : 1 GB
No. of processor : 2 Network cards : 02:00.0 Ethernet controller : Qualcomm Atheros Attansic L2 Fast Ethernet (rev a0)
System company : JOOYONTECH
Architecture : x86_64
OS flavour : Linux Mint 17 Qiana

Router Company : ipTIME N604R
Maximum speed : 160Mbps (LAN to WAN)

Tomcio

unread,
Oct 7, 2014, 8:10:03 PM10/7/14
to
Hello,

been here since years, but very low number of questions so I'm so happy
to see new thread here ;)

I'm not boewulf cluster admin but..

I belive Your network is slow because this routing device is internally
limited to low throughput, You need Gigabit-capable routing device (read
this as: Gigabit routing switch supporting layer 3).
I suspect all Your class rooms are in different IP subnets?
If yes then the routing device is bottleneck..

Even modern Core i5 based linux router will create additional delay
times in routing when it comes to cluster network, of course under high
load at high throughput (what clustering requires). So simple, raw
networking is best option.

Example capable switch models that can do routing are:
Cisco SG500X
Cisco Catalyst series
Dell PowerConnect 7000 series (or lower)
Of course, they could be too expensive so just look around on switch
that can do IP routing. There are even 8-port Gigabit Managed switches
available on market.

If Your sys admin isn't blocking other network subnets You could set
static IP on all Your cluster nodes to different network subnet (let's
say.. 192.168.99.1-192.168.99.10 with netmask 255.255.255.0) and check
if they see each other (probably this isn't possible) and use one of
them to provide Internet access (master node?) with 2 Gigabit cards.

USB Gigabit cards are most up to ~480Mbit in throughput since they're
USB 2.0. It's half of needed performance and! be aware of missing Linux
drivers.. ;)
I see all Your hardware is Gigabit, so I belive You need Layer 3 Gigabit
switch. Also, please check are Your NICs supporting offload functions,
then can help with high network load.
I personally these in /etc/rc.local
echo "Setting offload functions on Intel PRO/1000 NICs..."
ethtool -K eth1 rx on tx on sg on gso on gro on tso on
ethtool -K eth2 rx on tx on sg on gso on gro on tso on

And there are tips:
http://cs.boisestate.edu/~amit/research/beowulf/beowulf-setup.pdf
See the section: 1.2 Networking Hardware

Hope this helps.

Cheers,
TooMeeK




W dniu 2014-10-07 05:50, suresh kannan pisze:
--
To UNSUBSCRIBE, email to debian-beow...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listm...@lists.debian.org
Archive: https://lists.debian.org/54346B82...@o2.pl

suresh kannan

unread,
Oct 9, 2014, 11:40:03 PM10/9/14
to
Hi,

The link you have provided me was very helpful. I haven't prepared myself before starting clusters. Few years back i did clustering with three system using PVM for different application i.e protein ligand docking. With three systems it took some 2 hours to finish a particular library screening. Since, it is few hours i was happy and haven't checked the speed of the clusters. I thought to myself i learned a clustering technique. Therefore, this time i started blindly before preparing clustering for different applications which takes months to complete a specific job. Therefore, now i learned depending upon the purpose of clustering requirement of things will be different.

I learned that speed of the network is a biggest bottleneck for clustering especially for our need. In our lab I have found a 1000Mbs switch which presently reduces some time compared to 100Mbs however it is not efficient. Still it takes a lot of time. I assume we need 10gigabit switch. I was not aware of the price these "Gigabit routing switch supporting layer 3". It seems to me that even 8-port 10 gigabit switch cost approx. 800$. I am still hesitant to ask my mentor for the 10 gigabit switch. Since, I dont have no experience i had a thought what if i am wrong somewhere although gigabit switch have to work. Now, I am looking for 10 gigabit switch in near by labs so that i will connect and check whether it sufficiently efficient for our job and then i can ask my mentor with some confidence.

i also learned from someone we can use cluster OS specialized designed for this clustering purpose for instance Pelican cluster http://pareto.uab.es/mcreel/PelicanHPC/. I will install that cluster os and check the performance too. I am reading materials to increase the speed and performance of cluster in terms of hardware as well as the software's. I will tune our application according to our need.

And, also in our lab room we have dual boot systems for 8 people (windows with either linux mint, ubuntu, fedora). In given time those systems will be either in windows or linux environment depending upon our work. Daily 6 hours (nights) and on Sundays our system will be idle. I am not capable to use those free computer timing for our advantage due to my skill and also time. I guess we can use these computer time if we have gigabit switch with networking skills. If someone did similar stuffs please document either in your blogs or in email lists so that users like me might get benefited.

Although, it took time i learned a lot during this clustering. Thank you for your time.


regards
Suresh




--
To UNSUBSCRIBE, email to debian-beowulf-REQUEST@lists.debian.org

Tomcio

unread,
Oct 10, 2014, 2:00:03 PM10/10/14
to
Hello again,

Well.. with 10Gbit.. it's an overkill :D with this hardware lab.
Imagine that good 48-port 10Gigabit switch cost around 25-30k US $ (in
words: thousands). You'll also need 10Gigabit cards (let's say 500 US $
each). I didn't saw 8-port 10GbE switch anywhere, if You found it please
- let me know! :D

Shouldnt You stay at Gigabit? but with good performance switch.
It means high values of:
- switching capacity (in Mpps)
- switching bus (in Gbps)
and low value of:
- switching latency (fractions of ns)

Less expensive way is.. Infiniband network. It usualy starts at 10Gbit
speeds! There are Infiniband adapters and switches available on Ebay (I
own two of them) on PCI-Express x4 interface. Some, not all, support
TCP/IP networking as usual NICs.
This interface is specially designed for clusters due very low latency
times and low CPU usage at high speeds. But all cluster nodes have to be
close to each other since cables ARE expensive :)
They also support more equipment than networking (for example: storage).
But on beginning make sure that Infiniband card is well supported under
Linux, my isn't.

Regards,
TooMeeK




W dniu 2014-10-10 05:34, suresh kannan pisze:
> http://cs.boisestate.edu/~__amit/research/beowulf/beowulf-__setup.pdf <http://cs.boisestate.edu/~amit/research/beowulf/beowulf-setup.pdf>
> See the section: 1.2 Networking Hardware
>
> Hope this helps.
>
> Cheers,
> TooMeeK
>
>
>
>
> W dniu 2014-10-07 05:50, suresh kannan pisze:
>
> I am an Indian student in suwon, korea. I built a Beowulf cluster
> (system information below) with four systems in our lab for our
> simulation work with the help of good tutorials. In those
> tutorials they
> have mentioned all the system should have static ip addresses.
> Unfortunately, in all our labs we have been provided with dynamic ip
> address[5 ips for 15 members in three separate labs]. I have
> requested
> four more ip's from our university system admin. Due to the language
> problems, i conveyed the requirement through my korean lab mate
> and i
> dont know the reason why he denied us the static ip. So i found
> another
> way to skip this procedure
> http://www.reddit.com/r/__linuxquestions/comments/__2gubad/why_static_ip_address___is_necessary_for_linux/
> <http://www.reddit.com/r/linuxquestions/comments/2gubad/why_static_ip_address_is_necessary_for_linux/>.
>
>
> Someone suggested to use a router (one static ip) and set static
> ip for
> the four computers through a router. I did that and it worked.
> However,
> the cluster is very slow. For instance If i submit my simulation
> job in
> a single computer [4 core processor], it takes 2 months to
> complete a
> specific job. Although, if i connect 4 systems it shows it take
> 6 months
> to complete the same job. It is actually using 10 core processor
> [3,3,2,2-100% each]. I used TOP command to see how much
> processor the
> head and other nodes are using. I have used openMPI to do
> parallel the
> systems. I am using GROMACS (Parallelization based on MPI has
> been part
> of this software). I followed a parallel configuration for the
> Gromacs
> with the help of this tutorial
> http://flakrat.blogspot.kr/__2013/04/how-to-compile-__gromacs-461-with-openmpi.html
> <http://flakrat.blogspot.kr/2013/04/how-to-compile-gromacs-461-with-openmpi.html>.
> After reading few posts
> http://www.reddit.com/r/__linuxquestions/comments/__2gbgbg/what_would_be_the_best___linux_distro_for_folding/
> To UNSUBSCRIBE, email to debian-beowulf-REQUEST@lists.__debian.org
> <mailto:debian-beow...@lists.debian.org>
> with a subject of "unsubscribe". Trouble? Contact
> listm...@lists.debian.org <mailto:listm...@lists.debian.org>
> Archive: https://lists.debian.org/__54346B8...@o2.pl
> <https://lists.debian.org/54346B82...@o2.pl>
>
>


--
To UNSUBSCRIBE, email to debian-beow...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listm...@lists.debian.org
Archive: https://lists.debian.org/54381E69...@o2.pl

Rogerio Bastos

unread,
Oct 13, 2014, 3:50:02 PM10/13/14
to
First of all, why do you think network is the bottleneck?
Be sure of this before spend money with network equipments!

On 2014-10-07 00:50, suresh kannan wrote:
> I am an Indian student in suwon, korea. I built a Beowulf cluster
> (system information below) with four systems in our lab for our
> simulation work with the help of good tutorials. In those tutorials
> they have mentioned all the system should have static ip addresses.
> Unfortunately, in all our labs we have been provided with dynamic ip
> address[5 ips for 15 members in three separate labs]. I have requested
> four more ip's from our university system admin. Due to the language
> problems, i conveyed the requirement through my korean lab mate and i
> dont know the reason why he denied us the static ip. So i found
> another way to skip this procedure
> http://www.reddit.com/r/linuxquestions/comments/2gubad/why_static_ip_address_is_necessary_for_linux/
> [1].
>
> Someone suggested to use a router (one static ip) and set static ip
> for the four computers through a router. I did that and it worked.
> However, the cluster is very slow. For instance If i submit my
> simulation job in a single computer [4 core processor], it takes 2
> months to complete a specific job. Although, if i connect 4 systems it
> shows it take 6 months to complete the same job. It is actually using
> 10 core processor [3,3,2,2-100% each]. I used TOP command to see how
> much processor the head and other nodes are using. I have used openMPI
> to do parallel the systems. I am using GROMACS (Parallelization based
> on MPI has been part of this software). I followed a parallel
> configuration for the Gromacs with the help of this tutorial
> http://flakrat.blogspot.kr/2013/04/how-to-compile-gromacs-461-with-openmpi.html
> [2]. After reading few posts
> http://www.reddit.com/r/linuxquestions/comments/2gbgbg/what_would_be_the_best_linux_distro_for_folding/
> [3] i suspected the network router might be an issue.
> Links:
> ------
> [1]
> http://www.reddit.com/r/linuxquestions/comments/2gubad/why_static_ip_address_is_necessary_for_linux/
> [2]
> http://flakrat.blogspot.kr/2013/04/how-to-compile-gromacs-461-with-openmpi.html
> [3]
> http://www.reddit.com/r/linuxquestions/comments/2gbgbg/what_would_be_the_best_linux_distro_for_folding/

--

My email was sent by May First/People Link
https://mayfirst.org


--
To UNSUBSCRIBE, email to debian-beow...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listm...@lists.debian.org
Archive: https://lists.debian.org/40cfbd085315754e...@mail.mayfirst.org

suresh kannan

unread,
Oct 14, 2014, 10:50:02 AM10/14/14
to
Hi,


>>>First of all, why do you think network is the bottleneck?

Thank you for your concern.

If i run a specific job in a single system (without parallel) it showed three months (aprox) time in a quad core processor.  When i did parallel using four systems it showed October 2015. I was using 100 Mpbs (ip time router). I have conformed that all the systems uses the processors [using "top" command]. After that TooMeeK point out i have to use "Gigabit routing switch supporting layer 3". I dont have that switch in our lab. However, we had Netgear GS608 1000 Mpbs switch it reduced the time (Feb 2015). Moreover, I read the link http://cs.boisestate.edu/~amit/research/beowulf/beowulf-setup.pdf suggested by TooMeek. They have also suggested that Network switch is important.  Also, I saw few videos in youtube about layer 3 switch capabilities. Therefore, I am convinced that "Gigabit routing switch supporting layer 3" will solve this issue.

regards
Suresh








--
To UNSUBSCRIBE, email to debian-beowulf-REQUEST@lists.debian.org

with a subject of "unsubscribe". Trouble? Contact listm...@lists.debian.org
Archive: https://lists.debian.org/40cfbd085315754eb7f15af3a3dfc94...@mail.mayfirst.org


Tomcio

unread,
Oct 14, 2014, 1:20:02 PM10/14/14
to
Well..

It depends.
Based on what You're writting I assumed gateway (router) is routing
traffic between Your machines or You're using 100Mbit switch for Your
Gigabit-available nodes or it's just too slow switch.

When this situation occurs:
1 system - 3 months
4 systems 100Mbit - 1 year (4 times longer than single node)
4 systems 1000Mbit - 4 months (1,33 times longer than single node)
Then there is still something misconfigured or Your nodes aren't
optimized to work in cluster or Your switch is overloaded.

The task should take less than 3 months, let's say 1 month or less in
optimized cluster.

You could check network load on Your switch, but this model isn't even
web managed. Also, it's still relatively low performance (16 Gbps
switching capacity, 1,48Mpps per port, 15us switching delay).
So on Your servers there are available programs to measure network usage
like: ntop, dstat or iptraf (the last one not always show the correct
results) to help troubleshooting network performance.

Regards,
TooMeeK

W dniu 2014-10-14 16:42, suresh kannan pisze:
> Hi,
>
> >>>First of all, why do you think network is the bottleneck?
>
> Thank you for your concern.
>
> If i run a specific job in a single system (without parallel) it showed
> three months (aprox) time in a quad core processor. When i did parallel
> using four systems it showed October 2015. I was using 100 Mpbs (ip time
> router). I have conformed that all the systems uses the processors
> [using "top" command]. After that TooMeeK point out i have to use
> "Gigabit routing switch supporting layer 3". I dont have that switch in
> our lab. However, we had Netgear GS608 1000 Mpbs switch it reduced the
> time (Feb 2015). Moreover, I read the link
> http://cs.boisestate.edu/~amit/research/beowulf/beowulf-setup.pdf
> <http://cs.boisestate.edu/%7Eamit/research/beowulf/beowulf-setup.pdf>
> suggested by TooMeek. They have also suggested that Network switch is
> important. Also, I saw few videos in youtube about layer 3 switch
> capabilities. Therefore, I am convinced that "Gigabit routing switch
> supporting layer 3" will solve this issue.
>
> regards
> Suresh
>
>
>


--
To UNSUBSCRIBE, email to debian-beow...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listm...@lists.debian.org
Archive: https://lists.debian.org/543D59B2...@o2.pl

Rogerio Bastos

unread,
Oct 20, 2014, 1:00:01 PM10/20/14
to
Is the code running in the 'single system' the same running in parallel
?
Is the software using network efficiently ?
What about network traffic ? It's high ?
Why do you need a layer 3 ? Are cluster's nodes in different network ?
I'm not sure if your problem is network bottleneck.

On 2014-10-14 11:42, suresh kannan wrote:
> Hi,
>
>>>> First of all, why do you think network is the bottleneck?
>
> Thank you for your concern.
>
> If i run a specific job in a single system (without parallel) it
> showed three months (aprox) time in a quad core processor. When i did
> parallel using four systems it showed October 2015. I was using 100
> Mpbs (ip time router). I have conformed that all the systems uses the
> processors [using "top" command]. After that TooMeeK point out i have
> to use "Gigabit routing switch supporting layer 3". I dont have that
> switch in our lab. However, we had Netgear GS608 1000 Mpbs switch it
> reduced the time (Feb 2015). Moreover, I read the link
> http://cs.boisestate.edu/~amit/research/beowulf/beowulf-setup.pdf [6]
>>> [1]
>>> [2]
>>>
>>
> http://flakrat.blogspot.kr/2013/04/how-to-compile-gromacs-461-with-openmpi.html
>>> [2]
>>> [3]
>>>
>>
> http://www.reddit.com/r/linuxquestions/comments/2gbgbg/what_would_be_the_best_linux_distro_for_folding/
>>> [3]
>>
>> --
>>
>> My email was sent by May First/People Link
>> https://mayfirst.org [4]
>>
>> --
>> To UNSUBSCRIBE, email to debian-beow...@lists.debian.org
>> with a subject of "unsubscribe". Trouble? Contact
>> listm...@lists.debian.org
>> Archive:
>>
> https://lists.debian.org/40cfbd085315754e...@mail.mayfirst.org
>> [5]
>
>
>
> Links:
> ------
> [1]
> http://www.reddit.com/r/linuxquestions/comments/2gubad/why_static_ip_address_is_necessary_for_linux/
> [2]
> http://flakrat.blogspot.kr/2013/04/how-to-compile-gromacs-461-with-openmpi.html
> [3]
> http://www.reddit.com/r/linuxquestions/comments/2gbgbg/what_would_be_the_best_linux_distro_for_folding/
> [4] https://mayfirst.org
> [5]
> https://lists.debian.org/40cfbd085315754e...@mail.mayfirst.org
> [6] http://cs.boisestate.edu/%7Eamit/research/beowulf/beowulf-setup.pdf

--

My email was sent by May First/People Link
https://mayfirst.org


--
To UNSUBSCRIBE, email to debian-beow...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listm...@lists.debian.org
Archive: https://lists.debian.org/89187b3fef245c9f...@mail.mayfirst.org

suresh kannan

unread,
Oct 21, 2014, 8:50:04 AM10/21/14
to
Hello,

>>>Is the code running in the 'single system' the same running in parallel ?

If i install the software only in the head node (/usr/local/) mpirun says the software is not available in other nodes and the job terminates automatically. So, I have installed Gromacs (the software for simulation) only in head node (/home/mpiuser/)  and it automatically installed in all other nodes in /home/mpiuser/ (due to ssh without password, i guess so). 

>>>Is the software using network efficiently ?

I have searched google but i am unable to find out. How can i find out that the software uses network efficiently. May be can I find that in openmpi manual?

>>>What about network traffic ? It's high ?

i will use "ntop' to find out. Presently i am running a job it will finish in a week and i will start troubleshooting the cluster things and update the thread .

>>>Why do you need a layer 3 ? Are cluster's nodes in different network ?

I learned from few documentations Master node must have two network cards for this type of cluster but we dont have in our systems. It seems from the discussions even the USB network card is slow. I dont know how to add additional network cards to the system. I am running clusters in the same network (only four systems using a single router).

It seems to me from the discussions that there are three problems

1) we dont have two network cards systems.

2) The router i was using (iptime 100Mbps). Although, the speed of those jobs increased still it is not efficient when i used 1000Mbps switch compare to a single system.

3) I have used this tutorial  http://flakrat.blogspot.kr/2013/04/how-to-compile-gromacs-461-with-openmpi.html (for gromacs) to install openmpi, gromacs. Only difference is i have used gcc compiler instead of intel compiler. That is the only difference (even they say gcc is make faster code than intel compilers) . I have read the config (log files for open mpi, gromacs, fftw libraries etc..) for errors but i found the installations of the softwares were quite ok. Even if i did some misconfiguration how do i find out that other than reading the config log files.  Is there are any other way?

I have observed one more thing  when installing gromacs. In three system i have CPU_ACCELERATION (AVX_256) and one system has SSE2. http://www.gromacs.org/Documentation/Acceleration_and_parallelization#CPU_acceleration.3a_SSE.2c_AVX.2c_etc. However, even if i run mpi job in three systems (configure with AVX_256) it does not increase the speed comparatively.


Thank you for your time,

regards





--
To UNSUBSCRIBE, email to debian-beowulf-REQUEST@lists.debian.org

with a subject of "unsubscribe". Trouble? Contact
listm...@lists.debian.org
Archive:


--

My email was sent by May First/People Link
https://mayfirst.org


--
To UNSUBSCRIBE, email to debian-beowulf-REQUEST@lists.debian.org

with a subject of "unsubscribe". Trouble? Contact listm...@lists.debian.org

Rogerio Bastos

unread,
Oct 21, 2014, 9:10:02 AM10/21/14
to
On 2014-10-21 09:42, suresh kannan wrote:
> Hello,
>
>>>> Is the code running in the 'single system' the same running in
> parallel ?
>
> If i install the software only in the head node (/usr/local/) mpirun
> says the software is not available in other nodes and the job
> terminates automatically. So, I have installed Gromacs (the software
> for simulation) only in head node (/home/mpiuser/) and it
> automatically installed in all other nodes in /home/mpiuser/ (due to
> ssh without password, i guess so).

If you are using MPI, you can run the core in parallel within the same
machine (1 process for each cpu core).
I think this is a god test to compare parallel running time with
non-parallel because there isn't network involved.


>>>> Is the software using network efficiently ?
>
> I have searched google but i am unable to find out. How can i find out
> that the software uses network efficiently. May be can I find that in
> openmpi manual?

Debugging...but if you don't write the code this is hard.


>>>> Why do you need a layer 3 ? Are cluster's nodes in different
> network ?
>
> I learned from few documentations Master node must have two network
> cards for this type of cluster but we dont have in our systems. It
> seems from the discussions even the USB network card is slow. I dont
> know how to add additional network cards to the system. I am running
> clusters in the same network (only four systems using a single
> router).

You don't need a layer 3 switch (or a router) to connect head node with
others nodes. A gigabit switch is enough.
Normally the head node have two network cards (A and B) because one (A)
connect it with cluster's nodes and the other (B) connect the head node
with users' network. So you access the head node by ip address of
network card B and the ip address of network card A is used only for
internal cluster communication. But this setup is not mandatory, is the
one most used.


> It seems to me from the discussions that there are three problems
>
> 1) we dont have two network cards systems.
No problem.


> 2) The router i was using (iptime 100Mbps). Although, the speed of
> those jobs increased still it is not efficient when i used 1000Mbps
> switch compare to a single system.
If possible, use gigabit switch with gigabit network cards.


> 3) I have used this tutorial
> http://flakrat.blogspot.kr/2013/04/how-to-compile-gromacs-461-with-openmpi.html
> [2] (for gromacs) to install openmpi, gromacs. Only difference is i
> have used gcc compiler instead of intel compiler. That is the only
> difference (even they say gcc is make faster code than intel
> compilers) . I have read the config (log files for open mpi, gromacs,
> fftw libraries etc..) for errors but i found the installations of the
> softwares were quite ok. Even if i did some misconfiguration how do i
> find out that other than reading the config log files. Is there are
> any other way?
I have never used this software, so I don't know. But if you can compile
without errors with gcc it's ok. I personally prefer gcc because is open
source.


--

My email was sent by May First/People Link
https://mayfirst.org


--
To UNSUBSCRIBE, email to debian-beow...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listm...@lists.debian.org
Archive: https://lists.debian.org/91dee58225ee4719...@mail.mayfirst.org
0 new messages