Notes

1) Server issues with JElizabeth had nothing to do with server issues and have to do with code - the codebase needs a clean rewrite but as resources don't exist for that we have to cure symptoms not the root cause of the white pages)

2) HempSmart was moved to another server - this resulted in an issue that is being discussed with Immedion as the server is not indicating it is within our whitelisted range. I am in discussions with them on how to re-configure web09 so that its not using an IP address not registered to us when making external calls. In the mean time the IP address was white-listed the same day it was reported by Estefanie to me - I did try calling Caren to let her know and could not reach her so replied to email to her directly and cc'ing PayQuicker.

3) HempSmart Reports stopped functioning for less than 1 hour as I reverted their system to the old report server immediately on being informed. David confirmed with me all reports were then behaving. I do not know if he informed the client or confirmed with her. This was added to the list of things to check for when adding a new server as we now understand the root cause and it has been fixed but also added to the checklist when adding a new server for MPP (reporting was not used by JElizabeth)

4) Someone, I have to assume it would be a CSR or Salesperson, informed HempSmart that they could mix and match autoship when we clearly stated that no one should be using the new autoship if:
a) they had associated autoship
b) had any autoships with the old configuration

5) Puratae synching - is files not being synched between servers. This is also resulting when designers are not paying attention and doing updates on both servers in the correct directories consistently. Prices on external tools appear to have dropped significantly since we last reviewed them so am testing applications on servers to identify which we should purchase.

recommendations:

A) allocate 2 senior developers to trace all white-page, page slowdowns, out of memory etc are resolved - they still exist but we don't have clients screaming about them as no one is hitting the system like Jelizabeth was.

B) keep checklist up-to-date and keep all senior developers in the loop on the checklist - Reports issue was something Richard most likely knew when setting up a reports server on a new system but that information was not relayed to anyone.

C) Make sure no one tells a client about the new autoship functionality unless they have first discussed it with development staff to determine what it will take to convert existing settings over. And/or properly determine if a pre-existing client could start with the new system. (Will provide you with list of clients with/without autoship).

D) In fact no new feature should be given to any client without further testing on every variant of the system and/or (and this is more important) that the behavior and actions of new features need to be considered by everyone as we all have clients we specifically work with and have more knowledge about those clients usage of the system so that it can all be brought to bear. When you (or I) send out something for input - everyone must provide input that week (preferabbly that day but that's not likely to happen). What tends to occur is that people see a request, plan to respond to it later, and then fail to respond at all until its brought up in a staff meeting. We also have a tendency to not fully scope a task and then we have scope creep (case in point Party Plans)

E) Purchase synching software to keep:
a) skins
b) client uploaded files/images
synchronized -- and look into whether the same mechanism can/should be used for code deployments. The latter being done manually and not real time.
I am currently testing https://www.puresync.net for full synch capability would want the Professional Version on each server so that if synch failed in one direction it would be caught in the opposite direction. Approximately $100 total

F) With all accounts > $x/month there needs to be a standing call with the client on a weekly basis. That call would be no longer than 15 minutes. And for those over $y/month we give them access to Skype channel that all CSRs are listening on for anything urgent (this is already in place with DLG, HyFlux). Also with these accounts, the salesperson should be following up by phone or email (or both) at least every 2 weeks.