Slow apps and unpredictable speed are the needles that often irritate customers. The irritation lasts for 10-15 seconds only; but the post effect of that lasts in memory of customers. So a small 10 seconds delay gets magnified by the past bad memories, and everytime the customer sees slowness, the irritation levels go on an exponential scale.

The tech teams always love new stuff; keep on building features, whether user likes it or not; have fun. But the ops or devops teams are the ones who bare the brunt of end users. When the ops says that we need to fine tune the apps, the apps team will say better to dig your own backyard, ie. infrastructure and then come to us. Very tricky and easy way to put the ball on the other side.

Ops team gets some tools, app team gets some tools, each one proves not guilty. But the end user is still not happy. Especially when you are a data center guy, and an app team has hosted apps on you, OMG! I can imagine your frustration. When you press for the invoice payment, that is when all these questions will start hitting you. On 13th of Mar, 11.05pm to 11.50pm, we got 89 support tickets and all point to slowness of the apps; as infra provider, you missed to meet the SLAs; hence we deduct 35% from the invoice! Not a good spot to be in at all.

How long this Tom and Jerry game can go on?

If you are really serious, as an infra provider or a data center guy, you need to have monitoring of all layers and assets of your DC. Hardware assets, software assets, 3rd party apps assets, network assets etc. all must be monitored. If you do not have proper data about your own IT assets and their performance, you will end up wasting time in proving not guilty.

Life is great. Do not waste your days in proving not guilty.