by mgimena | Jan 23, 2017 | News & Events
Here Comes The Cavalry To The Rescue
- While we are updating your monthly valuations data to stretch to the last 13 months, it is not advisable for you to use that feature.
- Indeed, there is a very, very minor risk that the data for the month you need is just being onboarded at the time you view the screen and it may not be 100 % complete at the time.
- In the meantime, you can use the following spreadsheet to assist you in calculating the number of units you had in stock per SKU.
- We anticipate that the SellerLegend Monthly Valuation data update will be completed in a couple of days, so this spreadsheet is only useful during that period, or when you need to get data that is more than 13 months old.
- It is OK to share it this spreadsheet with people who do not subscribe to SellerLegend.
- https://docs.google.com/spreadsheets/d/1-wthD6g_cygYC6f8XY_AsIX4BQ4N084UkXLnYO1sFDo/
PS. The SKU/FNSKU in the sample sheet are fake 🙂
by mgimena | Jan 21, 2017 | News & Events
There’s Good News And There’s Bad News
The bad news is that, for a few days, you MAY experience a degradation in your online response times.
The good news is that this is for TWO good causes:
- We have decided to collect your Monthly Valuation Reports for the past 13 months, rather than from just – at best – April 2016.
- The rationale for that is that this report is useful for tax year end and your company tax year end can be at any time in the past 12 months.
- This now requires us retrospectively to go back to Amazon and request and process those reports for a longer period of time, thus adding an extra burden to the servers while we do that.
- Getting us closer to a ‘Hey, Amazon, gimme my money back’ facility, we are ready to deploy a new screen which will show you all your reimbursements
- Since you can request a reimbursement for up to 18 months ago, we need to request all that data from Amazon as well and process it.
These two processes will add some load to our servers. We looked into temporarily getting an additional server for this, but we concluded that this would not alleviate matters as we still would end up hammering the database, which is where the bottleneck is.
We apologize if this causes you inconvenience, but, remember, this is only temporary.
by mgimena | Jan 7, 2017 | News & Events
We have just completed the last step in our annual disaster recovery test and we are happy to report that it was very successful.
The test simulates a situation where we have lost all of our 6 servers, lost the system code and lost our database due to a catastrophic event.
We managed to recreate a working environment from scratch and tested that we had no loss of data in less than 6 hours.
We have proved that our backups are properly created, that they are readable and can be restored onto a fresh set of servers to resume operations.
Full disclosure: we are not resilient to the loss of Amazon’s server infrastructure in US West (Oregon). At this stage of our growth, we cannot yet afford a multi-site infrastructure.
