Week 1: Quantum Circuit Simulation

[nyameko]

Week 1: Revisiting the Student Cluster Competition Selection Round

• Topics: Prepare your head node and compute node virtual machines, as well as an introduction to Quantum Computing
• Tasks:
  • Redeploy your cluster infrastructure on Sebowa
  • Familiarize yourself with GitHub Project Management tools
  • Write a simple Python program using Qiskit to create a quantum circuit with 1 qubit, apply a Hadamard gate, and measure the result. Pay attention to Tutorial 3: Section 10.iii & Tutorial 4: Section 3.ii

[mpofukelvintafadzwa]

Hi, my name is Kelvin Mpofu. I will be your mentor for this projects. Feel free to ask any questions.

[mmhashatsi]

Hey team @thandiwe-01 @auretja @Fortune678 @MavishDevOps

For any project related communication, please write a comment under the issue that you are working on e.g. if you have a problem with a task in week 1, you would write a comment in the issue under week 1 and similar for all other weeks.

Please @your mentor when you need him to attend to something so that he can get a notification and he will do the same.

Please make sure that everybody in your team and the mentor has accepted the git invite and are added to the project.

Good Luck!

[thandiwe-01]

Hi Kelvin hope you are well, just checked my Git now, sorry for the delay we had Exams on Friday and Saturday. Will notify the team ASAP, to check their git notifications.

[MavishDevOps]

I am currently busy with examinations as my leader thandiwe has mentioned .I will be free from following thursday. I humbly apologise for not being present

[MavishDevOps]

I am deeply grateful to the Chpc team, to ms Mabatho and team . Thank you with the bottom of my heart

[mpofukelvintafadzwa]

@thandiwe-01 Your exams are definitely more important. All the best with the exams. Let me know how you get along with it when you start.

[thandiwe-01]

@mpofukelvintafadzwa we are currently busy with Tut1, the HPL part is a bit challenging my teammate Sipho is currently busy with it and will inform you on his progress with it or when he is unable to solve the issue. He is just trying to figure it out a few times and if his unable to he will definitely contact you. I will be busy with the Quantum program on Thursday as I will than be free. But no worries from the 2nd of November till the 28 we have many free days amongst us approximately 21 days, the team will be fully on board more and get through the set weekly tasks with out much interruption from the exams. November has more free days than October, so we are good.

[MavishDevOps]

I will assist my team as I finish my exam on Thursday 14 november 2024 following week thursday .

[thandiwe-01]

@mpofukelvintafadzwa Good day Kelvin is it possible to have weekly online MS teams meetings to discuss the tasks, the days I would suggest looking at everyone's schedule would be Friday and Sunday around 4:30 pm or later. Either just one day or two depending on the amount of time we would like to discuss the tasks and challenges we experience. We will still use git for further questions during the week.

[mpofukelvintafadzwa]

Hi @thandiwe-01 yes we can do that. Sunday is preferable for me, 4:30pm is fine as well. Thanks.

[mpofukelvintafadzwa]

@thandiwe-01 My email is kmpofu@csir.co.za. if you can, please do setup the meeting. Many thanks.

[thandiwe-01]

Hi @nyameko we have encountered an error when trying to compile HPL. This is the error we get We have edited the Make.unisasct file as instructed

[thandiwe-01]

@nyameko Sorry the error is not clear

[thandiwe-01]

@nyameko With loging into nodes issue I am able to do so on the Sebowa.Openstack console but I am unable to type in commands on the console. On Windows powershell I still get this error.

[nyameko]

Hi @nyameko we have encountered an error when trying to compile HPL. This is the error we get We have edited the Make.unisasct file as instructed

Hi @thandiwe-01 , you've correctly identified that it's a PATH issue... You need to correctly set the LAlib = <PATH-TO-WHERE-YOUR-SYSTEM-INSTALLED-ATLAS>

[nyameko]

@nyameko Sorry the error is not clear

Also, you are strongly advised NOT to run the benchmarks as root, use the ubuntu user instead until such time as you've configured your users with sudo privileges.

[nyameko]

@nyameko With loging into nodes issue I am able to do so on the Sebowa.Openstack console but I am unable to type in commands on the console. On Windows powershell I still get this error.

I suspect that you have incorrectly associated the SSH key you created to your head node. Please review this section from the tutorials.

You need to make sure that for your ubuntu user on your head node VM, you've correctly copied over the contents of ~/.ssh/id_ed25519.pub from WORKSTATION(s), to your /home/ubuntu/.ssh/authorized_keys file on your head node VM.

[Fortune678]

Hi @nyameko we have encountered an error when trying to compile HPL. This is the error we get We have edited the Make.unisasct file as instructed

Hi @thandiwe-01 , you've correctly identified that it's a PATH issue... You need to correctly set the LAlib = <PATH-TO-WHERE-YOUR-SYSTEM-INSTALLED-ATLAS>

Hi @nyameko i believe it no longer a path issue, but it not there. When you look into the ATLAS directory which should contain , libsatlas.so and libtatlas.so but those two libraries are not there. We used a command like [ find / -name " libsatlas.so" ] but nothing showed up. To install the two libraries ( libsatlas.so and libtatlas.so ) you use this commands [ sudo apt install openmpi libatlas-base-dev ] but these two packages will still not exist.

[nyameko]

Hi @thandiwe-01,

Please accept my apologies, the Makefile should be configured as follows:

• Ubuntu-based systems

   ARCH               = <TEAM_NAME>
  
   MPdir              = /usr/lib/x86_64-linux-gnu/openmpi
  
   LAdir              = /usr/lib/x86_64-linux-gnu/atlas/
   LAlib              = $(LAdir)/libblas.so $(LAdir)/liblapack.so
  
   CC                 = mpicc
  
   LINKER             = mpicc

I've added you as a reviewer to the PR I've just submitted. Please test the above changes, then go to the PR to add your first comment / review 😄

[nyameko]

Hi Team,

You need to urgently finish off Tutorials 1 & 2. Then you will skip all of Tutorial 3, except the Qiskit Quantum Volume section. Similarly you'll skip Tutorial 4, and only do Section 3.ii.

After that you can close off this week and move onto Weeks 2 & 3 of your project, and complete them with @mpofukelvintafadzwa on Jupyter Labs.

[thandiwe-01]

Thank you we are on it we are currently multitasking between the virtual cluster and our physical cluster, We have gone through week 2 quantum circuits with @mpofukelvintafadzwa this past Saturday. Hopefully by Thursday we will send screenshots of our week 1 or 2 Quantum circuit activities.

[thandiwe-01]

*This is on my PC will soon run this program from the cluster. Step 1: Step 2: step 3: step 4: step 5: step 6: output

We are just trying out the activities first on our laptops.

[thandiwe-01]

Hi @nyameko and @mpofukelvintafadzwa we are stuck with this error, thought qiskit and other necessary installations were installed however this error.

[mpofukelvintafadzwa]

@thandiwe-01 try installing qiskit again. That error means it's not installed or maybe you installed it in the wrong location.

[thandiwe-01]

Hi @mpofukelvintafadzwasetting up a virtual environment in UBUNTU can be tedious . Is this the expected outcome after running the quantum program, in tut 3 I did not parametrize the variables. Should we run the week 1 quantum circuit activity in our cluster also?🤷‍♀️

[nyameko]

It's the second time I'm advising you guys to stop working as root, use the ubuntu until you've setup other sudo users.

Virtual environments seem tedious, but make things much easier for you because different python projects using different libraries will not contaminate your working tree...

Just remember that everytime you log out / start again, if you are not using a terminal multiplexer / session manager, you'll need to active you venv again.

I think we should move onto week two.

[thandiwe-01]

Hi @nyameko we redid tut3- 10 iii, and Eish I am the guilty one of having the habit of working as root, the other teammates not so much I noticed this past Saturday. Setting up the virtual environment is not a tedious task, we just fell into the rabbit whole of trouble shooting, I did notice I just needed to insert 3 after the word python, but different websites gave me a long way about creating a virtual python environment. Our Mentor notified us that the CHPC-scc git, steps were sufficient, we just needed the 3 as we have already realized it. Plus we learnt about parametrizing variables and running them on a backend this past Saturday.

Unfortunately still unable to ssh into the nodes through our laptops terminal for the VM. I think we might be missing a step, VNC works fine though.

[nyameko]

Excellent work on identify your problems, resolving them and moving onto next steps!

Which machine is BlackBull? And which machine is LAPTOP-38JLJCCA...?

You need to make sure you create a public and private key pair for EACH machine that you want to connect to your VM's...

Use the three levels of the -v switch to investigate your errors further... i.e. ssh -i <PATH> <user>@<server> -vvv.

I think you guys should follow Tut 2, and configure Ansible to create a user for each of your group members.

Some distros, ssh clients (i.e. the one on BlackBull-1) may restrict connections from user . I.e. from the Fingerprint the key pair was created as root...

Good Luck!

[thandiwe-01]

Thank you will follow the given instructions, BlackBull and LAPTOP-38JLJCCA... those are our individual laptops.

The images below are for Tut 4-3ii -some part of the code

[nyameko]

Hi @thandiwe-01 ,

I think we need to close this issue especially if you've sorted out your ssh problems....?

[MavishDevOps]

Tutorial 2 Ansible and Wireguard @nyameko @mmhashatsi

I had experienced issues doing zerotier as my university proxy blocked internet access to install it i will retry zerotier I have completed tutorial 2 last friday sections 11 -13 allocated by my team lead thandiwe and co lead sipho

I had done most of the tasks on our physical cluster we designed with our mentor on the university campus

The following are screenshots of all the tasks I completed for tutorial 2 on my physical cluster , note I had experienced glitches while doing the tutorial I tried to deliver the best output ,critical valuable feedback is welcome on how I can best complete these tasks

On section 11 I and my colleague sipho worked on ansible under the guidance of my mentor mr auret Inventory file for our cluster

testfile.txt

Tasks for checking if ansible inventory reached nodes in cluster

I tried , im new to ansible I do not know what these warnings suggest ,any help would do 😊

I have tried to do wireguard according to my understanding. Im unsure why I cannot ping 10.0.0.2 and 10.0.0.3 I have tried to figure out the peer_c_public key I generated another sshh key on wireguard

I have experienced some issues I am learning and any valuable feedback is welcome

With regards to peer_c_public key which key are the instructions referring to

Thank you all

[MavishDevOps]

Re:Tutorial 2 This is for zerotier when I try to do the next i get this error any valuable assistance is welcome

[nyameko]

Hi @thandiwe-01 ,

I think we need to close this issue especially if you've sorted out your ssh problems....?