CPUs vs C2 CPUs vs Committed CPUs vs Committed C2 CPUs

363 views
Skip to first unread message

Jafar Muhammed

unread,
Jun 22, 2020, 11:10:38 AM6/22/20
to gce-discussion
Hello Community,

I wish to use the C2 series.

But I need to have a quota upgrade.

When I tried to request for Quotas, I am a little confused.

I see, CPUs, C2 CPUs, Committed CPUs and Committed C2 CPUs.

Again, I wish to use a C2 instance in any Asian zones with a Commitment.

Here, which one I need to request for a quota upgrade?

Is Committed C2 CPUs enough or I need to have CPUs > C2 CPUs and then Committed 

C2 CPUs altogether as a parent-child relationship style?


pralove

unread,
Jun 22, 2020, 10:25:05 PM6/22/20
to gce-discussion
The total number of virtual CPUs across all of your VM instances in a region is the CPU quota.

Please follow this link to learn about requesting additional resource quota. 

Compute-optimized machine types which include the C2 machine types are ideal for compute-intensive workloads.

C2 machine types offer up to 3.8 GHz sustained all-core turbo and built on the latest generation Intel Scalable Processors (Cascade Lake).

Please follow this link to find about the available regions and zones for C2 maching types. 

Purchasing a committed use contract is purchasing compute resource (vCPUs, memory, GPUs, and local SSDs) at a discounted price in return for committing to paying for those resources for 1 year or 3 years.

The resources for which you can purchase commitments is found in this link.

Found a stackover flow thread on how to increase Committed CPUs quota.

I hope the links provided above is helpful for you.



Found a stackover flow thread on how to increase Committed CPUs quota. I hope this might be helpful.[4]

Jafar Muhammed

unread,
Jul 3, 2020, 2:58:14 PM7/3/20
to gce-discussion
Thank you, @pralove

I failed to get an increase in quota despite being tried multiple times.

I have contacted GCP's Sales Team through https://cloud.google.com/contact a couple of times.

I have received zero response to my contact submission as well.

My CC is on file, and I have balance in my GCP account from the $300 credit promo.

Moving on 🤦‍♂️

Benjamin Daniel

unread,
Jul 3, 2020, 10:53:02 PM7/3/20
to gce-discussion
Good luck getting any answers. I've sent multiple emails, called sales and opened multiple tickets and the best answer they could give me was, "Wait a month or two and try again".

The sales rep said the quota team changed requirements for committed use but won't tell anyone, sales reps included, what they changed them to.

We're all just shouting into the void here until Google decides we deserve to know. Anyone care to tell me how this is superior to AWS? I'm becoming less convinced every day.

Jafar Muhammed

unread,
Jul 6, 2020, 9:12:15 AM7/6/20
to gce-discussion
Hi Benjamin,

I have tested GCP and AWS.

I found that when I check the speed test from https://developers.google.com/speed/pagespeed/insights/, TTFB of GCP version is relatively faster.
I guess due to the EBS volumes networking; the TTFB took slightly longer in AWS. However, the TTFB is steady and within the limit.

But in general, AWS is as fast as GCP.

However, I couldn't test with actual specs as from GCP. I was limited to N1 instances. So, testing further was not fair.

Even if I have received the response from GCP's sales team after a week, I would still be happy.

No matter how large is the service provider's pocket, I don't trust and rely on my production infra on a space where I have no clue what will happen.

I was expecting at least an acknowledge from the sales team; they failed me =D

Another case, being the largest ESP, Google doesn't open Port 25, and that's a critical decision factor to me.

See https://googlecloudplatform.uservoice.com/forums/302595-compute-engine/suggestions/12422808-please-unblock-port-25-allow-outbound-mail-connec

After this terrible experience, I have moved to AWS, and I am so happy with their customer support, network and infrastructure.

Digil (Google Cloud Platform Support)

unread,
Jul 6, 2020, 2:16:27 PM7/6/20
to gce-discussion
Hello

Usually, the sales team should contact you either via email or call once you submit your necessary request. Providing a valid justification for your business concern is equally important as it drives the pivotal role for your request. The Quota management team of Google Cloud Platform will look for several factors before approving a resource quota for a project. Your long billing history, usage history, business needs, and proper justification are some of them from the lists of factors.


That's being said, if your Google Cloud Platform account is having a TAM(Technical Account Manager) or CE (Customer Engineer), I would strongly recommend you get in touch with them and explain the urgency of using this product. However, before contacting them, kindly ensure that you have the history of contacting sales and its details.


Google Compute Engine allows outbound connection on all ports but not on the port 25. The port 25 is blocked in-order to avoid the risk of abuse. I am not sure how critically important is your business need for port 25, but you could refer this help center article which describes various other options for sending mail from a virtual machine instance and provides general recommendations on how to set up your instances to send email.



Benjamin Daniel

unread,
Jul 6, 2020, 4:56:48 PM7/6/20
to gce-discussion
My concern is why this is even necessary. I can hop on AWS right now an buy the equivilant reserved instance without any justification or red tape. Why does Google make this so infuriatingly difficult?

I've contacted sales, multiple times, and so far all they've been able to tell me is that the quota team won't approve my request and I should wait longer. How long? They're not sure. Maybe try again in a month or two? Which helps me absolutely none when I need the service now. 

I should have known better when I signed up for GCS instead of going with AWS, this is such typical Google behaviour. Provide a "robust" service option with abysmal customer service that eskews interaction with actual people to help solve issues and when that inevitably fails they simply shrugs and says "Maybe someday". I used to have an unhealthy amount of trust in Google and their technical prowess and overall corporate structure. Now all I have is skepticism and mistrust that anything they provide, even paid services, will provide even the bare minimum in customer service. 

Bye Google, thanks for letting me know where your priorities are.

Justin Reiners

unread,
Jul 6, 2020, 5:44:19 PM7/6/20
to Benjamin Daniel, gce-discussion
Once you are established on GCP, you should have no issues spinning up anything, I've been a heavy GCP user for years, I don't have to ask for a thing anymore. I'm sure it's just for new accounts. I've got a couple 64 core machines for an API some N1, some C2, just started using the C2 instances last month.

I don't think I've ever requested an increase for anything 
Justin Reiners
Reiners Cloud Consulting LLC
I'm happy to be your local Omaha distributor for 3CX and SimpleHelp! Contact me for info.


--
© 2018 Google Inc. 1600 Amphitheatre Parkway, Mountain View, CA 94043
 
Email preferences: You received this email because you signed up for the Google Compute Engine Discussion Google Group (gce-dis...@googlegroups.com) to participate in discussions with other members of the Google Compute Engine community and the Google Compute Engine Team.
---
You received this message because you are subscribed to the Google Groups "gce-discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gce-discussio...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gce-discussion/f850b9cd-76d5-4bf5-ab5e-bc0c2cde698dn%40googlegroups.com.

dick.r...@gmail.com

unread,
Jul 21, 2020, 12:53:04 PM7/21/20
to gce-discussion
I'll note the $300 student credit fails its ostensible goal of converting
users while they're young. As deep learning initiates, most undergrads (and
many graduate students) seek GPU acceleration which remains off-limits at
GCE. Google Colab, on the other hand, is doing an enviable job of capturing
university mindshare by offering free TPU and GPU time -- just in an awkward
way (I'd rather run offline tensorflow scripts on GCE than negotiate
jupyter kernels).
Reply all
Reply to author
Forward
0 new messages