Our CTO, Darwin Lee, recently blogged about “Doing it Right”, and said, “There can be a lot of misinformation about who does what and how”. He was responding to assertions that TransVault solutions only connect to legacy archive source data via the archive in question’s API, which is nonsense to anyone who’s worked with our products. TransVault is, in fact, the ONLY archive migration vendor that offers its Partners and end-customers choice over which source connection methodology they want to employ. We’ve known the nuance around API vs Direct extraction for years…look, we even blogged about it in 2015…worth a read for anyone seeking to cast aspersions on their competitors.
The logic goes that connection to source data via API is the reliable way to go but can be slow, and that going Direct is generally associated with performance benefits, although it might compromise data integrity in some cases (if done poorly). The ideal, of course, is that you have both extraction methods available to you when scoping a migration project. There will be situations when one method is more appropriate than the other, depending on the data at hand, as well as myriad other factors like the status of the archive’s API health and/or load when you need to extract from it. Some archives that TransVault support extraction from don’t publish an API, so in those cases we have no choice but to go direct to the data.
TransVault knows all about this dichotomy, and that’s why we have a hybrid extraction connector for Enterprise Vault and the choice of API or Direct connections (where possible) for the other archive sources we deal with day in, day out.
Extraction via the source archive’s API, by definition, gives you the highest fidelity because, after all, what better way to retrieve an item than via the technology that stored it in the first place. That’s why we offer API connections. However, API has it’s downsides – you’re relying on the source system, which means it needs to be up and running – it may be slow, it may be busy – and actually retrieving terabytes of data over a short period can cause reliability problems in the platform itself. These platforms weren’t specified or designed for a scenario where all data is going to be retrieved as quickly as possible.
With Direct, you bypass all the issues related to the source archive and just deal with the native data itself. We have lots of expertise and over 2000 projects under our belt that prove that we know what we’re doing when it comes to dealing directly with source platform data.
Having Direct extraction as your only answer equals failure.
Direct cannot always be the answer for an EV migration. There are ways in which Enterprise Vault can store data which is impossible to access directly. Customer can often use the “secondary storage” feature to move older data to other locations, like into Netbackup, or even to the Cloud, and such data cannot be accessed at the direct level. So that’s why any EV migration solution worth its salt has to support API as well.
You can see then, that the ONLY right answer is to offer both approaches, giving you the best of both worlds depending on your needs.
The truth of the matter is that, no software vendor should tell you one of these methods is the only way to go for your migration, and that other vendors in the market use the opposite approach, so you should beware – that would be leaving out the nuance of the situation and not giving the customer a clear view of the pros and cons of each approach. Ideally then the technology should step in and make life easier on the customer’s choice of approach, rather than dictating the only way to go.
Darwin wrapped up our last post by saying, “please don’t leave yourself needing a hat”. Let’s finish this one by saying…you have a choice of haircuts, so make sure you pick a hairdresser who gives you the full story before they start chopping bits off.