What is our primary use case?
I am currently working with Riverbed for replication between PR to DR for synchronization purposes, and for WAN optimization between replication from one data center to another data center. The main use case of Steelhead is for replication and branch connectivity for WAN optimization purposes. The goal is to achieve WAN optimization. For example, if I have a 500 GB link and use this product for WAN traffic over the 500 GB, it reduces traffic by 70 to 90%. The 500 GB traffic would reduce to around 100 GB. The main purpose of Steelhead is reducing bandwidth from 500 GB to around 100 GB. This reduction helps in lowering the cost of bandwidth that was previously paid to ISP service providers. Riverbed solution also claims it reduces latency and response times.
What is most valuable?
The logs in Steelhead are fantastic. There is a deep level of logs such as top 10 docker and top 50 docker. I can check from the top 50 docker which type of application is optimizing well and which type of traffic is not optimized. Steelhead provides logs and percentage levels, which is good. I can see any type of log report, report security, different types of report availability, but it is not customizable. Overall, this is a good feature.
Steelhead provides real-time optimization with graphs and tables on real-time optimization, informing us packet by packet including port, source IP, destination IP, and destination port number. This reporting and real-time monitoring is fantastic, although I faced a problem in Oracle.
Steelhead mainly saves the money that needs to be paid to ISPs. My actual traffic is around 2 GB, but I purchased a link from the ISP for only around 700 to 800 MB. All the prices that need to be paid monthly to the ISP are saved, which is a significant saving for my company after using Riverbed.
What needs improvement?
In my experience with Steelhead, monitoring tools, logs, and live monitoring features can be improved. These are fantastic now, but a few additional options can be added. Oracle performance is a main concern. Steelhead is very expensive, and the vendor should work on handling Oracle type of traffic better.
For how long have I used the solution?
I have been working with Steelhead for the last three years.
What do I think about the stability of the solution?
Steelhead is a stable device. I can say it is stable due to its significant features. For instance, even if there is a power failure, Riverbed still works. If the power is unplugged from Riverbed, it goes into fail-to-wire mode.
What do I think about the scalability of the solution?
Steelhead is not too scalable due to its licensing scenario. I have a model that provides only two licensing options. When I asked my vendor partner to increase the bandwidth, they told me the box is in full utilization license. Therefore, it is not scalable, as there are only two options: 300 or 600 MB. If a customer has 1 GB traffic, they will only be able to utilize 600 MB of traffic with Steelhead.
How are customer service and support?
I have no issues with technical support from Riverbed, and I would rate them eight out of ten.
How would you rate customer service and support?
Which solution did I use previously and why did I switch?
In Pakistan, Riverbed is only focused on WAN optimization. For NPM, there are many other solutions working such as SolarWinds and others. Steelhead solution is the only Riverbed solution available in Pakistan for WAN optimization.
How was the initial setup?
The setup process for Steelhead is somewhat complex because the behavior of Riverbed is different from other networking devices such as firewalls, routers, and switches. There is one in-path, but the in-path has two different interfaces, which can confuse beginners during deployment. The deployment process is a little difficult, but after deployment, operation is smooth with no issues. Cable connectivity is acceptable.
Which other solutions did I evaluate?
I have considered multiple competitors, but I have not worked with any of them, such as Cisco, which also provides a WAN optimization solution, so I cannot comment on their products.
What other advice do I have?
Application performance may improve with Steelhead, but it varies application to application. Some applications, such as Oracle traffic, are not working well, but other traffic performs fine. The traffic is passing, but Oracle DB traffic experiences lag and slowness due to Steelhead, while other applications perform fine.
There are very few improvements. I cannot say anything definitively, but it may have reduced page load time by 5 to 10%. This is not a significant improvement. With Oracle, my Oracle DB traffic was very slow.
Overall, I am satisfied with Steelhead. I would rate the price for the product as seven to eight due to it being high. I would rate Steelhead with a final mark of six to seven, with six being a better option in my opinion.
Which deployment model are you using for this solution?
On-premises