It is fantastic for disaster recovery. We've used it in DR tests for years and it just works. This is the primary reason I can think of why a company would use this feature set.
SharePlex is best suited for replicating relatively small but steady streams of data, instead of huge amounts of data surge. Keep in mind that it relies on database logs (redo logs in Oracle) to replicate data in the form of "messages", and therefore it is mandatory to put the database in forced logging mode - that may be a prohibiting factor in some situations.
Disaster Recovery and Business Continuity are strong areas where NetApp SnapMirror exceeds by providing multiple recovery options without the need for additional software or degrading network performance.
The volume size on the destination partner always has to be bigger than the source. This can get annoying. Not sure if they have made this automatic in future releases.
When you break a SnapMirror to use the destination volume you have to redirect all requests to the new volume. This isn't really a con but it is something you need to think about.
Documentation: There are some areas where the documentation is weak or dated. In particular, running SharePlex on RAC is a more involved process than the implementation guides would have you believe. When reaching out for assistance, the answer from Dell (a few years ago) was to engage their consulting team.
Built in scripts/agents: SharePlex comes with a handful of monitoring and management scripts that can be run as cron jobs. Unfortunately, these scripts are primitive and rigid, requiring the user to hard-code values into the script body and limiting their usefulness when you're running multiple replication streams. Some of the methods and dependencies are heavy. We ended up writing custom scripts that do what we need but it would be helpful if there were greater functionality and flexibility out of the box.
Compared to other tools, there is not as wide (or at least as active) a user presence. If you're having difficulties you may have to engage a consultant since it's less likely that you'll find an answer in a forum or blog post.
Very well versed and a good product to deal with downtime to recover from the failure is pretty good. Also, NetApp's engagement on Incident calls is very professional and humble. I would give a 100% rating to this product as I have been working on this for the past 5 years. It is stable, fast and highly reliable
It just works 99% of the time. Even if there's a network outage, jobs just pick up after the outage is over. The only reason it's not a 10 is the weakness of the GUI interface for configuration. For me, all configuration has to be done at the command line.
NetApp support has always been great. With the NetApp Auto-support system/web portal, the visibility into your NetApp storage is incredible. There have been times when I've been contacted by them for an issue that I haven't even discovered yet. For example, I received a call 20 minutes after a drive failed in our filer. I hadn't checked my email at that time and therefore didn't see the alert.
Support is very responsive. They take ownership of the problems and see them through to the finish. When a bug is found they work towards developing testing and making the fix available to us to solve the problem.
We originally had an EMC storage device. We had quickly outgrown the device and looked for alternatives. We were drawn to NetApp because of the great software solution surrounding it. We had been using a combination of backup solutions previously combining the Dell EMC box with Backup Exec and a tape library. NetApp replaced all of that nicely. No more tapes. Seamless replication no only locally but also to a remote location.
SharePlex was a clear winner in comparison to its competitors. It beats GoldenGate on price, providing equivalent features and performance at a fraction of the price and without the need to license costly Oracle features at the database level. DB Visit was a strong contender for us as well and it is an excellent product I would also recommend, but it did not have quite enough of a user and community presence and penetration for our management to feel comfortable with.
