<html><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space;" class=""><head><meta http-equiv="Content-Type" content="text/html charset=utf-8"><meta http-equiv="Content-Type" content="text/html charset=us-ascii"></head><div><meta http-equiv="Content-Type" content="text/html charset=utf-8" class=""><div class=""></div><div class=""><br class=""></div><div class="">For some perspective, here is the same graph of hourly connection failures plotted on a log scale vs overall data request traffic through the same host. There’s not really a way to predict when things will start falling apart from our end. All we see is the number of failures periodically jumps up during times of modest activity. These tend to happen about the same time that the rest of the cyverse apis see an increase in usage, but when compared to the overall cpu, memory, disk, io, and network utilization on the servers running these Data services in question, there is a strong indicator that gremlins are involved.</div><div class=""><br class=""></div><div class=""><a href="https://dl.dropboxusercontent.com/u/203259/irods_request_vs_errors_5-23-15_5_30_15-logscale.pdf" class="">https://dl.dropboxusercontent.com/u/203259/irods_request_vs_errors_5-23-15_5_30_15-logscale.pdf</a></div><div class=""></div></div></body></html>