Marketing and communication efforts need to be improved. Many in this world think Solaris is dead or dying. This idea has to be stopped and even reversed in order for Solaris to regain market share. Solaris is one of the best OSs out there today, and everyone seems to think it's going away. If Oracle spent more time informing people of what they're doing WITH Solaris vs laying off their developers then we would see a lot more people adopting this superior OS.

The tool is quite complex and difficult for anyone trying to use or study it. The complexity of learning the tool is an area of concern where improvements are required. Oracle Solaris' lighter and simpler version can be introduced. I feel that Oracle Solaris is a user-friendly tool because I am exposed to it.

Currently, it supports the DataFast file system. Sometimes we face some issues with the DataFast file system where there is some sort of slowness, which is intermittent. If it could support other file systems, it would be better.

I don't want to receive any updates on Oracle. The aforementioned detail can be considered for improvement.

When we switch over to Solaris it was not easy because we had some troubles with the performance. Solaris is from Oracle and you would expect that it would run flawlessly, but we had some issues in sizing the previous Linux environment to the Solaris environment.

Most of the product is still command-line, despite the fact that they've got a graphical user interface in some areas. For some reason, core administration is still done via command-line.

The manufacturer can put most of those command-line environments into classical use like other operating systems. With Solaris the administration part is through command-line which may be difficult for some people who may not be used to that way of working.

Currently, there are two variants, there's SPARC and there's x86. I would have wanted a scenario where they're all just one product.

I would have loved if the clustering data was a bit simpler. Currently, the clustering data is a product on its own. It would be great if there was higher availability data with that.

The scalability of the solution can be improved. 

The solution is pricey and can be improved by lowering the cost.

The primary drawback with this product is the lack of version updates. Despite my association with Oracle, I find it a concern that they are persisting with version 11.4 without transitioning to version 12. Updating to a new version would convey a sense of innovation and progress, even if they are incorporating numerous enhancements, fixes, and security measures.

Solaris' package management could be improved, especially in comparison to Linux.

I haven't had any big issues with the solution. Largely, I've been very happy with it. 

Sometimes we get stuck for security reasons. There's an issue with either the filing or use management. However, largely it just comes down to a lack of experience and over time, users get the hang of it. 

It would be helpful if the solution offered backend management. In the 11.4 version, Oracle added a management console. It would be great if we maybe had a user management tool to go with it.

More monitoring tools could be included in the product.

I would like to see a graphical user interface (GUI) improved.

Moreover, there is room for improvement in documentation. 

Oracle Solaris can improve by supporting all the recent features that are in the market from other competitors.

It is not easy to use. It doesn't have a user-friendly interface. It should be easy to use. We are planning to move from Solaris to Linux because Linux is more flexible and user-friendly. Its installation should also be easier. 

Solaris also needs specific hardware to work well, which is another reason why we are moving to Linux. It should be more flexible in terms of hardware. It should have better integration with other hardware platforms.

There are some areas that could use some improvement. As with Solaris 10, you can install Solaris 11 on SPARC and x86 systems, but the number of non-Oracle x86 systems certified up to this point is less that with the previous version. In spite of that, you can still install Solaris 11 on a varied number of systems as 'bare metal' or you can resort to virtualization via many of the softwares available for that in the market. The certification of third-party hardware is usually a lengthy process and requires a lot of resources, so it would be understandable if this takes a long time.

This product is not as flexible as other similar solutions on the market today. Times have moved on and there are newer operating systems that are better to use and more compatible with current technology.

I have noticed very frequent HBA and NIC card failure in T5-1B or T4-1B blade servers. More stability is required. I have experience multiple instances where QUAD HBA+NIC Port, for T4-1B, T5-1B was failed. And we had some downtime, as replacing HBA card, potentially needs server to be powered-off. Though this was noticed in case of Blade servers only, not for T5-8 or T5-2 or T4-2 etc.

The product is very expensive. 

The Solaris code is open, and the documentation is accessible to all, not only to registered users. Also, the documentation does not support some solutions, and there are no other options.

I would like to see improvements in user adaptability, the poor UI, and in the packaging of the applications. Solaris should look like Linux and the end user should not be afraid of using it. It's way different than linux.

It should have POSIX compatibility with Linux.

Patching without downtime would be nice.

Update 08/2021: Live Paching of Kernel is now available. We applied IDRs successfully on several servers.

There is an issue where Solaris doesn't give the correct figures for memory use when checked.

They could also enable Oracle OEM for x86 architecture as well. It is currently applied only to risk processors. It could help reduce costs associated with x86 environments compared to risk environments.

Live migration of kernel zones would be the biggest improvement for us; the ability to migrate that from one hardware platform to another on the fly without downtime.

It may not be as widely adopted and lacks abundant resources and tutorials compared to other Linux systems. The challenge arises from the differences in commands and configurations compared to more common systems like Linux.

During my many years of use, I've suffered with small problems while implementing the advanced features of Solaris. They were always, however, resolved by referring to the old and legendary Sun Solve and, today, to the excellent Oracle documentation website (http://docs.oracle.com).

Oracle Solaris is a Unix-based operating system. People used to call it Slowaris because it was very slow. However, they have different CPU architectures. You can use Solaris also on a normal server x86. Additionally, they have their own CPU architecture, which is called Oracle SPARC architecture. I believe before Oracle was using Oracle Linux, they were using Oracle Solaris for their customers who are using Oracle databases. This was because it was more optimized for the hardware built for it. It has good performance for the database only. However, if you take it out of the Oracle applications, it will not do well compared to other operating systems, such as Linux or even Windows can have better performance.

I believe it's a great product and its latest versions are also really good. However, I believe Oracle is not utilising its full potential by restricting it best performance with Oracle hardware. Even though it can be run on SPARC, as well as Intel hardware, the problem lies with the way Oracle chooses to promote it. They are always saying that it performs best with Oracle hardware. They should understand current demand for open source and publish white papers for its performance on Intel hardware. And they must change their stratergy with Dell, HP and other blade server manufacturers and enable them to use Solaris and promote Solaris.

Also, they should promote Oracle Solaris with open source tools like HANA, Hadoop, Puppet, Chef, and Ansible. Meanwhile, they can continue to develop and promote their in-house competitive products as well.

To summarise, I feel the main issue lies with their promotion and sales strategies, and also their relations with competing hardware vendors and database/application vendors.

The total cost of ownership of SPARC vs. x86 will always be a consideration.

There is a lack of general availability of training outside of Oracle University.

We are Oracle partners so come from this from both sides of the fence.

Generally skills in Solaris (RISC) are on the decline whereas Linux is still on the rise. You can expect nearly every techie to have x86 windows and or Linux skills but Solaris is a niche skill that is getting harder to recruit for. Previously Solaris was for the enterprise and x86 was commodity however because of the improvement in hardware, x86 is generally a match for RISC in many cases but far cheaper.

If you were starting from scratch, most companies would opt for x86 because it’s cheaper and easier to find skills and training for. You can find any number of training courses for Linux online, classroom, book based, you tube etc etc. Solaris tends to be only supplied by Oracle University at premium rates.

It’s hard to say where I think Solaris can improve. I feel like Solaris is getting more and more into a niche. The problem Solaris has is, it's fighting against open source. Open source is taking more and more of the market. It's for free. Solaris you have to pay for. I think there is a problem. I feel it's a propriety system; not open to be modified by the community.

One cool feature with Oracle Linux, is the ability to patch without a reboot. Getting this working on Solaris would rock! With the new M7/S7 chips, better DTRACE visibility into the hardware acceleration offloading would be nice. Difficult to explain a server that is 90% idle but doing the workload of 4 Intel servers.

I think Oracle should also promote x86 architecture for Solaris, so that the same can be used in ESX and in cloud environments with an x86 variant. It would be a bonus for Oracle.

SPARC hardware is costly. Most businesses want to run their infrastructure environments - especially non-production environments - on x86 hardware, where customers can run heterogeneous OS platforms (Linux, Solaris and/or AIX). However, this is not possible with AIX at all (especially with Solaris). So, if Oracle improved x86 support in Solaris, it could promote Solaris x86 as having the same stability and reliability levels as Solaris SPARC servers. This would provide customers a reason to move their servers - which are currently migrating to Linux servers – to Solaris x86. And I am sure this will boost Solaris even further. I am well aware that Oracle is promoting Oracle Linux for the same reasons, but I think the same can be done in a better way for the existing Solaris OS.

Right now, we don't have any difficulties.

It's good. To me, it's better than other products. For example, Power AIX. I think Solaris better. 

Solaris is not easy to use. It needs better GUI, UI, and configuration tools.

Software availability.

The product is really good in and of itself. It does need more third-party support and applications.

What needs to be improved is the documentation. That's not an issue with the product per se, but the documentation lacks lots of things. It's very difficult to find related things. They are not referenced. When a document speaks about one topic, it almost never refers to related topics. That's a bad thing. Documentation speaks mostly about how to do things; it does not speak about why to do or not to do, when you have options. That's missing. Sun used to have such documentation. With Oracle, I don't see it.

I would probably like to see it open sourced once again, as was situation with Open Solaris.

Right now, I see less and less organizations using Solaris and, at least from my point of view, there is not much active development around it.

I’m working less and less with the OS it seems. Where I used to think, "I’d love them to improve this," I’ve heard that's exactly what they’ve done. Even the newest Oracle Mini Cluster only has a visual interface for deployment and management.

The OS always needed and still does need more of a visual interface. Not to take anything away from the command line - I love that - but for basic mass user community acceptance, there is a large Windows, under-30 user base that doesn’t know how to think when they don’t have a mouse to do things with.

• GUI or menu configuration support
• Management tool visualization

Since Solaris Volume Manager is the obvious choice for a shared filesystem, I would love to see improvements in SVM so that filesystems can be increased or migrated without downtime for the environment - in a similar way that ZFS is capable of.

Offering more GUI applications might help adoption of Oracle Solaris by professionals coming from the Windows “world”.

Also, adaptation of behavior of commands similar to those from Linux would add familiarity for Linux system administrators aiming to use Oracle Solaris as well. (E.g., in Linux, the ifconfig command with no parameters returns the output for all interfaces. In Oracle Solaris, it causes an error and usage help).

The product itself is great. I would like to see whether other companies start developing for Solaris.

Needs NTFS support and VMware compatibility. To install Solaris as a VMware virtual machine, I need to convert the VirtualBox image to a VMware image.

I would love to see improvements in SVM, so file systems could be increased or migrated without downtime to the environment, similar to what ZFS is capable of.

I would like it to be faster; sometimes it takes a while, including making a connection. It can be complicated at times, too.

In terms of monitoring, they have quite a bit right now, but it could be a little bit better, especially with all of the virtualized resources.