Leaving Oracle

I first came to Sun over 4 years ago for an internship in the Solaris kernel group.  I was excited to work with such a highly regarded group of talented engineers, and my experience that summer was no disappointment: I learned a lot and had a blast.

After college, I joined Sun’s Fishworks team.  Despite my previous experience at Sun, I didn’t really know what to expect here, and I certainly didn’t imagine then where I’d be now and how much I’d value the experiences of the last three years.  In that short time, I’ve worked with many bright people with diverse backgrounds, I’ve learned so much about so many topics (both technical and otherwise), and I’ve continued having fun contributing to a great product.

Now I’ve decided to try something different, this time outside Sun/Oracle.  My last day here is 9/24.  While I’m excited about the future, I’ll miss both the people and the work here.  I wish you all the best of luck, and thanks for everything!

SS7000 Software Updates

In this entry I’ll explain some of the underlying principles around software upgrade for the 7000 series.  Keep in mind that nearly all of this information are implementation details of the system and thus subject to change.

Entire system image

One of the fundamental design principles about SS7000 software updates is that all releases update the entire system no matter how small the underlying software change is. Releases never update individual components of the system separately. This sounds surprising to people familiar with traditional OS patches, but this model provides a critical guarantee: for a given release, the software components are identical across all systems running that release. This guarantee makes it possible to test every combination of software the system can run, which isn’t the case for traditional OS patches.

When operating systems allow users to apply separate patches to fix individual issues, different systems running the same release (obviously) may have different patches applied to different components. It’s impossible to test every combination of components before releasing each new patch, so engineers rely heavily on understanding the scope of the underlying software change (as it affects the rest of the system at several different versions) to know which combinations of patches may conflict with one other. In a complex system, this is very hard to get right. What’s worse is that it introduces unnecessary risk to the upgrade and patching process, making customers wary of upgrading, which results in more customers running older software. With a single system image, we can (and do) test every combination of component versions a customer can have.

This model has a few other consequences, some of which are more obvious than others:

  • Updates are complete and self-contained. There’s no chance for interaction between older and newer software, and there’s no chance for user error in applying illegal combinations of patches.
  • An update’s version number implicitly identifies the versions of all software components on the system. This is very helpful for customers, support, and engineering to exactly what it means to say a bug was fixed in release X or that a system is running release Y (without having to guess or specify the versions of dozens of smaller components).
  • Updates are cumulative; any version of the software has all bug fixes from all previous versions. This makes it easy to identify which releases have a given bug fixed. "Previous" here refers to the version order, not chronological order. For example, V1 and V3 may be in the field when a V2 is released that contains all fixes in V1 plus a few others, but not the fixes in V3. More below on when this happens.
  • Updates are effectively atomic. They either succeed or they fail, but the system is never left in some in-between state running partly old bits and partly new bits. This doesn’t follow immediately from having an entire system image, but doing it that way makes this possible.

Of course, it’s much easier to achieve this model in the appliance context (which constrains supported configurations and actions for exactly this kind of purpose) than on a general purpose operating system.

Types of updates

SS7000 software releases come in a few types characterized by the scope of the software changes contained in the update.

  • Major updates are the primary release vehicle. These are scheduled releases that deliver new features and the vast majority of bug fixes. Major updates generally include a complete sync with the underlying Solaris OS.
  • Minor updates are much smaller, scheduled releases that include a small number of valuable bug fixes. Bugs fixed in minor updates must have high impact or high likelihood of being experienced by customers and have a relatively low risk fix.
  • Micro updates are unscheduled releases issued to address significant issues. These would usually be potential data loss issues, pathological service interruptions (e.g., frequent reboots), or significant performance regressions.

Enterprise customers are often reluctant to apply patches and upgrades to working systems, since any software change carries some risk that it may introduce new problems (even if it only contains bug fixes). This breakdown allows customers to make risk management decisions about upgrading their systems based on the risk associated with each type of update. In particular, the scope of minor and micro releases is highly constrained to minimize risk.

For examples, four major software updates have been released: 2008.Q4, 2009.Q2, and 2009.Q3, and 2010.Q1. The first four of these have had several updates. We’ve also released a few micro updates.

Replication for disaster recovery

When designing a disaster recovery solution using remote replication, two important parameters are the recovery time objective (RTO) and the recovery point objective (RPO). For these purposes, a disaster is any event resulting in permanent data loss at the primary site which requires restoring service using data recovered from the disaster recovery (DR) site. The RTO is how soon service must be restored after a disaster. Designing a DR solution to meet a specified RTO is complex but essentially boils down to ensuring that the recovery plan can be executed within the allotted time. Usually this means keeping the plan simple, automating as much of it as possible, documenting it carefully, and testing it frequently. (It helps to use storage systems with built-in features for this.)

In this entry I want to talk about RPO.  RPO describes how recent the data recovered after the disaster must be (in other words, how much data loss is acceptable in the event of disaster). An RPO of 30 minutes means that the recovered data must include all changes up to 30 minutes before the disaster.  But how do businesses decide how much data loss is acceptable in the event of a disaster? The answer varies greatly from case to case. Photo storage for a small social networking site may have an RPO of a few hours; in the worst case, users who uploaded photos a few hours before the event will have to upload them again, which isn’t usually a big deal. A stock exchange, by contrast, may have a requirement for zero data loss: the disaster recovery site absolutely must have the most updated data because otherwise different systems may disagree about whose account owns a particular million dollars. Of course, most businesses are somewhere in between: it’s important that data be "pretty fresh," but some loss is acceptable.

Replication solutions used for disaster recovery typically fall into two buckets: synchronous and asynchronous. Synchronous replication means that clients making changes on disk at the primary site can’t proceed until that data also resides on disk at the disaster recovery site. For example, a database at the primary site won’t consider a transaction having completed until the underlying storage system indicates that the changed data has been stored on disk. If the storage system is configured for synchronous replication, that also means that the change is on disk at the DR site, too. If a disaster occurs at the primary site, there’s no data loss when the database is recovered from the DR site because no transactions were committed that weren’t also propagated to the DR site. So using synchronous replication it’s possible to implement a DR strategy with zero data loss in the event of disaster — at great cost, discussed below.

By contrast, asynchronous replication means that the storage system can acknowledge changes before they’ve been replicated to the DR site. In the database example, the database (still using synchronous i/o at the primary site) considers the transaction completed as long as the data is on disk at the primary site. If a disaster occurs at the primary site, the data that hasn’t been replicated will be lost. This sounds bad, but if your RPO is 30 minutes and the primary site replicates to the target every 10 minutes, for example, then you can still meet your DR requirements.

Synchronous replication

As I suggested above, synchronous replication comes at great cost, particularly in three areas:

  • availability: In order to truly guarantee no data loss, a synchronous replication system must not acknowledge writes to clients if it can’t also replicate that data to the DR site. If the DR site is unavailable, the system must either explicitly fail or just block writes until the problem is resolved, depending on what the application expects. Either way, the entire system now fails if any combination of the primary site, the DR site, or the network link between them fails, instead of just the primary site. If you’ve got 99% reliability in each of these components, the system’s probability of failure goes from 1% to almost 3% (1 – .993).
  • performance: In a system using synchronous replication, the latency of client writes includes the time to write to both storage systems plus the network latency between the sites. DR sites are typically located several miles or more from the primary site in case the disaster takes the form of a datacenter-wide power outage or an actual natural disaster. All things being equal, the farther the sites are apart, the higher the network latency between them. To make things concrete, consider a single threaded client that sees 300us write operations at the primary site (total time including network round trip plus time to write the data to local storage only, not the remote site). Add a synchronous replication target a few miles away with just 500us latency from the primary site and the operation now takes 800us, dropping IOPS from about 3330 to about 1250.
  • money: Of course, this is what it ultimately boils down to. Synchronous replication software alone can cost quite a bit, but you can also end up spending a lot to deal with the above availability and performance problems: clustered head nodes at both sites, redundant network hardware, and very low latency switches and network connection. You might buy some of this for asynchronous replication, too, but network latency is much less important than bandwidth for asynchronous replication so you often can  save on the network side.

Asynchronous replication

The above availability costs scale back as the RPO increases.  If the RPO is 24 hours and it only takes 1 hour to replicate a day’s changes, then you can sustain a 23-hour outage at the DR site or on the network link without impacting primary site availability at all.  The only performance cost of asynchronous replication is the added latency resulting from a system’s additional load for the replication software (which you’d also have to worry about with synchronous replication).  There’s no additional latency from the DR network link or the DR site as long as the system is able to keep up with sending updates.

The 7000 series provides two types of automatic asynchronous replication: scheduled and continuous. Scheduled replication sends discrete, snapshot-based updates at predefined hourly, daily, weekly, or monthly intervals. Continuous replication sends the same discrete, snapshot-based updates as frequently as possible.  The result is essentially a continuous stream of filesystem changes to the DR site.

Of course, there are tradeoffs to both of these approaches. In most cases, continuous replication minimizes data loss in the event of a disaster (i.e., it will achieve minimal RPO), since the system is replicating changes as fast as possible to the DR site without actually holding up production clients. However, if the RPO is a given parameter (as it often is), you can just as well choose a scheduled replication interval that will achieve that RPO, in which case using continuous replication doesn’t buy you anything.  In fact, it can hurt because continuous replication can result in transferring significantly more data than necessary. For example, if an application fills a 10GB scratch file and rewrites it over the next half hour, continuous replication will send 20GB, while half-hourly scheduled replication will only send 10GB.  If you’re paying for bandwidth, or if the capacity of a pair of systems is limited by the available replication bandwidth, these costs can add up.


Storage systems provide many options for configuring remote replication for disaster
recovery, including synchronous (zero data loss) and both continuous and scheduled asynchronous replication.  It’s easy to see these options and guess a reasonable strategy for minimizing data loss in the event of disaster, but it’s important that actual recovery point objectives be defined based on business needs and that those objectives drive the planning and deployment of the DR solution. Incidentally, some systems use a hybrid approach that uses synchronous replication when possible, but avoids the availability cost (described above) of that approach by falling back to continuous asynchronous replication if the DR link or or DR system fails. This seems at first glance a good compromise, but it’s unclear what problem this solves because the resulting system pays the performance and monetary costs of synchronous replication without actually guaranteeing zero data loss.

Replication in 2010.Q1

This post is long overdue since 2010.Q1 came out over a month ago now, but it’s better late than never. The bullet-point feature list for 2010.Q1 typically includes something like "improved remote replication", but what do we mean by that? The summary is vague because, well, it’s hard to summarize what we did concisely. Let’s break it down:

Improved stability. We’ve rewritten the replication management subsystem. Informed by the downfalls of its predecessor, the new design avoids large classes of problems that were customer pain points in older releases. The new implementation also keeps more of the relevant debugging data that allows us to drive new issues to root-cause faster and more reliably.

Enhanced management model. We’ve formalized the notion of packages, which were previously just "replicas" or "replicated projects". Older releases mandated that a given project could only be replicated to a given target once (at a time) and that only one copy of a project could exist on a particular target at a time. 2010.Q1 supports multiple actions for a given project and target, each one corresponding to an independent copy on the target called a "package." This allows administrators to replicate a fresh copy without destroying the one that’s already on the target.

Share-level replication. 2010.Q1 supports more fine-grained control of replication configuration, like leaving an individual share out of its project’s replication configuration or replicating a share by itself without the other shares in its project.

Optional SSL encryption for improved performance. Older releases always encrypt the data sent over the wire. 2010.Q1 still supports this, but also lets customers disable SSL encryption for significantly improved performance when the security of data on the wire isn’t so critical (as in many internal environments).

Bandwidth throttling. The system now supports limiting the bandwidth used by individual replication actions. With this, customers with limited network resources can keep replication from hogging the available bandwidth and starving the client data path.

Improved target-side management. Administrators can browse replicated projects and shares in the BUI and CLI just like local projects and shares. You can also view properties of these shares and even change them where appropriate. For example, the NFS export list can be customized on the target, which is important for disaster-recovery plans where the target will serve different clients in a different datacenter. Or you could enable stronger compression on the target, saving disk space at the expense of performance, which may be less important on a backup site.

Read-only view of replicated filesystems and snapshots. This is pretty self-explanatory. You can now export replicated filesystems read-only over NFS, CIFS, HTTP, FTP, etc., allowing you to verify the data, run incremental NDMP backups, or perform data analysis that’s too expensive to run on the primary system. You can also see and clone the non-replication snapshots.

Then there are lots of small improvements, like being able to disable replication globally, per-action, or per-package, which is very handy when trying it out or measuring performance. Check out the documentation (also much improved) for details.

Remote Replication Introduction

A bad copy?When we first announced the SS7000 series, we made available a simulator (a virtual machine image) so people could easily try out the new software. At a keynote session that evening, Bryan and Mike challenged audience members to be the first to set up remote replication between two simulators. They didn’t realize how quickly someone would take them up on that. Having worked on this feature, it was very satisfying to see it all come together in a new user’s easy experience setting up replication for the first time.

The product has come a long way in the short time since then. This week sees the release of 2010.Q1, the fourth major software update in just over a year. Each update has come packed with major features from online data migration to on-disk data deduplication. 2010.Q1 includes several significant enhancements (and bug fixes) to the remote replication feature. And while it was great to see one of our first users find it so easy to replicate an NFS share to another appliance, remote replication remains one of the most complex features of our management software.

The problem sounds simple enough: just copy this data to that system. But people use remote replication to solve many different business problems, and supporting each of these requires related features that together add significant complexity to the system. Examples include:

  • Backup. Disk-to-disk backup is the most obvious use of data replication. Customers need the ability to recover data in the event of data loss on the primary system, whether a result of system failure or administrative error.
  • Disaster recovery (DR). This sounds like backup, but it’s more than that: customers running business-critical services backed by a storage system need the ability to recover service quickly in the event of an outage of their primary system (be it a result of system failure or a datacenter-wide power outage or an actual disaster). Replication can be used to copy data to a secondary system off-site that can be configured to quickly take over service from the primary site in the event of an extended outage there. Of course, you also need a way to switch back to the primary site without copying all the data back.
  • Data distribution. Businesses spread across the globe often want a central store of documents that clients around the world can access quickly. They use replication to copy documents and other data from a central data center to many remote appliances, providing fast local caches for employees working far from the main office.
  • Workload distribution. Many customers replicate data to a second appliance to run analysis or batch jobs that are too expensive to run on the primary system without impacting the production workload.

These use cases inform the design requirements for any replication system:

  • Data replication must be configurable on some sort of schedule. We don’t just want one copy of the data on another system. We want an up-to-date copy. For example, data changes every day, and we want nightly backups. Or we have DR agreements that require restoring service using data no more than 10 minutes out-of-date. Some deployments wanting to maximize freshness of replicated data may want to replicate continuously (as frequently as possible). Very critical systems may even want to replicate synchronously (so that the primary system does not acknowledge client writes until they’re on stable storage on the DR site), though this has significant performance implications.
  • Data should only be replicated once. This one’s obvious, but important. When we update the copy, we don’t want to have to send an entire new copy of the data. This wastes potentially expensive network bandwidth and disk space. We only want to send the changes made since the previous update. This is also important when restoring primary service after a disaster-recovery event. In that case, we only want to copy the changes made while running on the secondary system back to the primary system.
  • Copies should be point-in-time consistent. Source data may always be changing, but with asynchronous replication, copies will usually be updated at discrete intervals. But at the very least, the copy should represent a snapshot of the data at a single point in time. (By contrast, a simple "cp" or "rsync" copy of an actively changing filesystem would result in a copy where each file’s state was copied at slightly different times, potentially resulting in inconsistencies in the copy that didn’t exist (and could not exist) in the source data.) This is particularly important for databases or other applications with complex persistent state. Traditional filesystem guarantees about their state after a crash make it possible to write applications that can recover from any point-in-time snapshot, but it’s much harder to write software that can recover from arbitrary inconsistencies introduced by sloppy copying.


  • Replication performance is important, but so is performance observability and control (e.g., throttling). Backup and DR operations can’t be allowed to significantly impact the performance of primary clients of the system, so administrators need to be able to see the impact of replication on system performance as well as limit system resources used for replication if this impact becomes too large.
  • Complete backup solutions should replicate data-related system configuration (like filesystem block size, quotas, or protocol sharing properties), since this needs to be recovered with a full restore, too. But some properties should be changeable in each copy. Backup copies may use higher compression levels, for example, because performance is less important than disk space on the backup system. DR sites may have different per-host NFS sharing restrictions because they’re in different data centers than the source system.
  • Management must be clear and simple. When you need to use your backup copy, whether to restore the original system or bring up a disaster recovery site, you want the process to be as simple as possible. Delays cost money, and missteps can lead to loss of the only good copy of your data.

That’s an overview of the design goals of our remote replication feature today. Some of these elements have been part of the product since the initial release, while others are new to 2010.Q1. The product will evolve as we see how people use the appliance to solve other business needs. Expect more details in the coming weeks and months.