What is our primary use case?
The main requirement from my clients to
Google Cloud is to reduce the on-prem footprint or to minimize the maintenance cost of the on-prem data center. Those who are looking for a scalable solution can go with
Google Cloud. The main key point is to reduce the footprint of the on-prem data center in terms of purchasing servers and maintenance renewals, which can mainly focus on the hardware. To reduce the hardware footprint, we can prefer Google Cloud.
Some functions or features I really value in Google Cloud, especially for the banking sector and BFSI, is that they would want to maintain their on-prem workloads segregated with multiple data centers, while still adopting part of their workload to Google Cloud. In that case, they want to replicate or keep a secondary copy in Google Cloud. Google has tie-ups with some ISV solutions, Independent Software Vendor companies. For example, on-prem we have Pure Storage, Pure, and now Dell (formerly Dell EMC). Pure, NetApp, and Dell have tie-ups and have created custom solutions. If I want to keep my primary copy on-prem and reduce the cost or minimize my hardware footprint, I can keep my secondary copy in Google Cloud. Google has tie-ups with NetApp, Pure, and Dell, and they have created customized ISV solutions. The secondary copy can be kept in Google Cloud while the primary copy remains on-prem, so we can replicate the data. Overall, the pricing and costing factors can be reduced. Instead of deploying the whole same setup on another data center, Google is providing these kinds of solutions. Even other vendors provide solutions, but when compared with Pure and their solutions, we do not have the equivalent in Azure or AWS. It may be in the development phase for those providers. I am giving one example for setting up and configuring a DR setup in Google Cloud. Instead of keeping everything on-prem, to set up an on-prem disaster recovery, we have to purchase switches, routers, and servers, among many other things. However, to minimize the cost, we can keep it in Google Cloud. That is one option. The second point is that for workloads, at least for development and non-prod environments, we can leverage Google Cloud. We can use Google Cloud for both prod and non-prod workloads and application workloads.
What is most valuable?
Google Cloud
Kubernetes support, especially
GKE, is really reliable and good mainly from a microservices point of view for any product development.
Google Kubernetes Engine is really reliable and it is good for product development because if you take some product, they might have to release multiple features on a release-by-release basis, adding multiple features. GKE is helping a lot with this. Earlier, we were completely in the monolithic approach because all the software and everything was clubbed and kept in one operating system. Now, we have segregated it. If I want a snapshot, I can use one node from the GKE. If I want some other feature in terms of replication-related features or OS-related features, everything I can use individual
Kubernetes nodes. Whenever I want to do some enhancement or updates specific to that specific service, we can easily do it. Google is good in that regard.
What needs improvement?
Regarding potential areas of improvement for Google Cloud, considering that they are providing 300 plus services, some services could be improved. For example, if I am facing some network troubleshooting issue in GKE in Google Cloud, I have to dig deeply and enable some special logs, such as flow logs. To enable that, it is chargeable. At least for basic services, they should give end-to-end support. In terms of end-to-end support, if I am purchasing one firewall from Google and using that firewall, all firewall-related services should be free. If I want to troubleshoot something deep and perform deep troubleshooting, you have to enable some additional logs. That is not that good. If I have to enable additional logs and want to store them into Google Cloud storage, which is similar to an
S3 bucket, it is again chargeable. Those kinds of things should be given as part of the package. If you take firewall, all firewall-related logs, activities, and diagnostics everything should come as one package. If you want to do some deep analysis, you have to enable these logs, and for that you have to store them in the cloud storage, which is again chargeable. We should not have those kinds of things.
Regarding additional costs, it can vary case by case. If I want to do deep analysis, I have to enable some logs, and for those logs, Google is charging. I am already paying for my virtual firewall from Google, and I am paying the cost for that. If I want to do some further troubleshooting and enable some low-level logs, they are charging for that. So these are the kinds of things I am talking about. To do very deep-level troubleshooting and enable additional logs, they will be charging. Not only Google, but all clouds are charging for this.
Functionality-wise, Google Cloud is fine. In terms of virtual machine deployment or VM instance, the functionality is good.
For how long have I used the solution?
I have been working with Google Cloud for two years.
What do I think about the stability of the solution?
Regarding the stability of the product, it is the best one. I can give it an eight out of ten.
What do I think about the scalability of the solution?
Scalability-wise, I will also give it an eight out of ten. Google Cloud is good for everything in terms of scalability.
How are customer service and support?
Technical support is also good. The support mark is based on whether you have standard or premium support, and based on that they will provide the support, which is fine.
Which solution did I use previously and why did I switch?
Currently, I am completely out of the
PowerProtect domain and now completely into the cloud with
Azure,
AWS, and Google Cloud. So I am no longer working with
PowerProtect, Data Domain, or Dell products.
I have quit DXC technology. I spent almost nine years with DXC. After that, I moved to LTI Mindtree, and now I am at Mindtree.
How was the initial setup?
The initial setup for Google Cloud is simple.
What other advice do I have?
The AutoML feature of Google Cloud is really good. Google is providing good AI features. We have been leveraging
ServiceNow integration through API. Whatever knowledge base articles are stored in
ServiceNow, through an ETL process (extract, transform, load), we have integrated with Google Engine through API. So if I want to get information, for example, my shift starts at 2:00 PM, and how many hours I am supposed to work, if I type in the front end through a chatbot or RIMA or anything, then we get the response. How we get the response is through Google Cloud. The Google Cloud storage talks to the Google search engine and AI search engine, and then it gets the data from ServiceNow, and then it responds to the user, such as "This is your shift time, 9 hours." We have been integrated with ServiceNow,
Confluence portal, and
SharePoint. It is really very helpful.
For security and IAM side, identity and access management in Google Cloud is good. In terms of roles and assignments, it comes under RBAC, which is role-based access control, and that is a very good feature. We can control at each service level and can restrict the control of the users. Only for developers, they need only read-only access for storage account or cloud storage. We can restrict the control at each service level. That is good.
Multi-cloud support in Google Cloud is good. Only a very few customers are using multi-cloud support. For example, a client has Azure Kubernetes Service and GKE. Both have been integrated using Google Anthos. Google Anthos is one of the services, and the main purpose is that we can deploy our workloads in AWS, for example, Kubernetes services and Kubernetes nodes. Similarly, I can deploy the same nodes on Google Cloud side as well with GKE. At any point in time, if my AWS whole region goes down, users can still access their workloads from GKE. To know that this goes down and the user is accessing from Google Cloud, the user means admin. Because Google is providing Anthos, there we can integrate with AWS as well as Google Cloud Kubernetes services. There we have the dashboard from where we can monitor. That is the best option. This option we do not have in AWS or Azure. Only Google has it with Anthos.
For Google BigQuery, we are using Google Cloud for Big Data purposes, and that is good. I have provided a solution for Big Data to only two to three clients on Google Cloud side. From BigQuery point of view, it is good. Google Cloud is good. I give this review an overall rating of eight out of ten.