On March 14th, 2019 at approximately 11:05 AM Pacific, one of our distributed filesystems experienced increased latency which resulted in read-only filesystems for less than 0.5% of paid Pantheon Sites. At 11:50, our engineers identified and implemented a remediation. By 12:00, the number of sites affected dropped by half, and by 12:20 all sites were operating as usual. We have since implemented an improvement to reduce load on our distributed filesystems, we have updated our internal documentation on how to mitigate this issue, and we are updating our alerts to be more sensitive to distributed filesystem latency.