Mission-critical applications like MS Exchange and SQL running in their virtual environment became progressively slower over time. Moreover, Exchange users began to timeout because of the overwhelming I/O demand that was exceeding the constraints of their infrastructure.
Telestream, LLC is an American privately held provider of software and hardware products for video capture, encoding, transcoding, and network based delivery.
Telestream virtualized 80% of their environment, and everything was running fine for about a year before things began to slow down – particularly with their most I/O intensive applications like MS Exchange, SQL, Jira, and more.
However, the biggest pain point for Telestream beyond sluggish application performance was the fact that Exchange users were timing out and getting disconnected from the email server. Telestream was pushing the I/O boundaries of their infrastructure, causing queue depth and latency spikes, which ultimately resulted in users timing out and getting disconnected from Exchange.
“We knew we had an I/O problem and had reached the IOPS limitation of our backend storage architecture under peak load. However, what we did not consider is how much of our storage performance was being dampened by small, fractured, random I/O characteristics due to Windows write inefficiencies that generate excessively small writes and reads. That was ultimately the root cause problem why our users were timing out from Exchange,” said Allie McLachlan, Systems Administrator, Telestream.
Telestream had been briefed on Condusiv’s V-locity® I/O reduction software for virtual environments a year prior, but had no need to boost application performance at the time. However, after Telestream began to experience slowdown in their virtual environment, they thought there was no harm in evaluating the software.
V-locity is “set and forget” software that runs transparently in the background on Windows servers and automatically offloads I/O from underlying storage, then streamlines the I/O traffic that remains. All of this is done with near-zero overhead to the CPU. First, V-locity eliminates excessively small, fractured writes and reads, and displaces them with large, clean contiguous writes so more payload is carried with every I/O operation. Second, V-locity establishes a tier-0 caching strategy by using idle, available DRAM to serve hot reads. Nothing has to be allocated for cache since V-locity dynamically adjusts to only what is otherwise unused.
With as little as 2GB of available memory, many customers serve as much as 50% of their read traffic. As a result, most V-locity customers experience at least 50% faster application performance, with many workloads getting much more, depending on the extent of Windows write inefficiencies and how much memory is available.
After evaluating the software, the timeout issues related to MS Exchange disappeared and Telestream saw an immediate 50% boost in application performance across their most I/O-intensive applications. Once the 30-day evaluation period expired, applications slowed and the Exchange timeouts began to happen all over again. It was clear to Telestream that the root cause problem was in fact Windows write inefficiencies that were overwhelming the infrastructure with excessively small writes and reads – a “death by a thousand cuts” scenario. For Telestream, the purchase of the software was considered a “no brainer.”
“As soon as we installed V-locity and saw the Exchange timeouts disappear, and users were no longer getting disconnected from Exchange, that made our purchase of V-locity I/O reduction software a ‘no brainer,’” said McLachlan.
McLachlan, continued, “The most eye-opening experience was finding out how badly our system performance was being taxed by small, fractured, random I/O from Windows write inefficiencies. We would have had no idea that was even an issue without trying V-locity to solve those root cause performance issues.”