3 ms·
I disagree. If an outage in the product your team managed was caused by a bug your team introduced 5 can point at several root causes and solutions that your te
by dtech 8y ago
I disagree. If an outage in the product your team managed was caused by a bug your team introduced 5 can point at several root causes and solutions that your team will be able to control and implement.
Management only comes into play when causes like "under-funding", "constantly changing requirements" or "too tight deadlines" are identified.
- vajrabum 8y agoManagement also has to support the idea of blamelessness and the costs of investigation and remediation and to participate in reviews. Without that support it may work some of the time but often won't.
- jdietrich 8y agoManagement have ultimate responsibility for your team - staffing, training, resources, workload and everything else. If one of those is wrong, it's their fault. The alternative is pure scapegoating. Blaming a team or an individual doesn't get to the root cause - why did that employee underperform, why was that team badly led, why did nobody identify or report these problems until it was too late? The answer might be as simple as "we hired a complete moron", but management have to take ownership of that error if they want to prevent similar errors in future.
- Darkphibre 8y agoReminds me I need to finish Extreme Ownership. Got about half way through, a really excellent philosophy.
- bluGill 8y agoManagement is still at fault for not putting in place enough quality to ensure that the bug didn't happen, or was found first. There are a number of ways to do this. They can spend more time and money testing the product before it goes to production. They can spend more money/time on automated testing infrastructure. They can spend time/money on failover to last good version. They can spend time/money on up front design. They can decide that the costs of any of the above is too high and so they will accept the risk. Depending on your exaction situation any of the above can be most cost effective (often a combination - in particular accepting risk eventually is required no matter how much you spend on something else).