>The ISP networking infrastructure that this product relies to has an IOPS issue which, apparently, could be resolved with some reengineering. This is something I reported a few years ago and provided a few hundreds of hours in an attempt to help them resolve it. I do not know the current status of that issue presently. I am no longer involved in it. However, this product depends from it. This usually happens overnight Eastern Time Zone when administrative procedures occur on the network. Sometimes, but very rarely, it happens during the day. This lasts between 30 to 60 seconds in average. Then, everything starts running normally. A few weeks ago, I did an adjustment in the product to limit the visual effect this could have on logged on users. However, as you can imagine, when the IOPS stop to respond on the entire network, no matter what we could try to fix it, it won't work as not even the OS could respond on all servers.

How about turning the problem around? Change the script in the page to detect the condition and just ignore it... until once it works and the user is back where he was. I mean the script which refreshes the toolbar and sidebars periodically.