Steemit Suggestion: Better User Notification When Steemit Fails - Such as During DDOS or Other Error

Several weeks ago Steemit was mostly unusable for several days and receiving ANY kind of update on the causes and likely ETA for a solution was almost impossible for most users. The communication from Dev team to end-users was mostly non existent. The result was predictably that chat rooms were filled with people needing to know what was occurring who then met with various theories about what was occurring that may or may not have been accurate.

Besides the fact that many users are put off from returning to a site that is consistently unreachable, the added concern of having wallets that are also unreachable adds an extra tension that causes unnecessary stress for man users. The situation is unnecessary, since, while we cannot always predict what errors and even attacks might occur, we CAN take design steps to ensure that users are always informed of the situation as they struggle to use the site when it is down/broken.

Currently...

---

When Steemit has an infrastructure problem that leads to pages being inaccessible, they simply come to the website and pages either load slowly or they may even see a standard NGINX server error of a 502, 503 or other variety. This situation demonstrates that not even the most basic error handling and user experience features have been implemented for the website.

Suggested Improvements

---

The most basic improvement to the situation might be to update the NGINX server config files to use custom error pages for each type of server error code, so that at least users feel that the system is not TOTALLY broken and that there is some semblance of awareness within the system that an error has occurred. Custom error page handling is simple to set up in NGINX and the error pages can contain scripts that notify devs of their presence and also that provide helpful links to users that might point them to external wiki or info pages that help them figure out what to do while the issues are resolved. Users might even be presented with a list of alternative Steem based websites that they can use to access the Steem blockchain and their wallets.

In the event that Ajax loads fail (content that is loaded internally in a page, after the page has initially loaded successfully) - the page might produce a popup error message or alternatively forward to a custom error page in the event that the error was of a kind that renders the page that was being accessed to be useless for the user (e.g. if the user is viewing a list of blogs and no blogs can be loaded).

A singular status page could be the target for such errors and the status page could be updated by the dev team with periodic, short statements of fact regarding the present state of the network, any known issues and the ETA for them being fixed. I understand that the dev team is prioritising other things ahead of this, but this is not the way to win over new users from other social networks who expect a certain level of interaction and information to be available as a bare minimum before they will consider the service to be reliable.

I understand that a custom DDOS protection layer is in place now at Steemit.com and maybe other error handling systems I am unaware of too. Since I don't know enough about the details to comment on them, I will simply suggest that a workable solution here is to ensure that the gateway server that is the main point of contact between the website and the rest of the internet is one that contains logic that can reliably present error/status notifications in the event that other servers go down or are unreliable. The gateway server can check for downstream errors and handle them appropriately in a way that at least provides a bare minimum of information to users about the errors.

---

Posted on Utopian.io - Rewarding Open Source Contributors

---