Saving months of compute time with a single Grafana query
1–10 of 54 posts
Re: Saving months of compute time with a single Grafana query
#2[flagged]
Re: Saving months of compute time with a single Grafana query
#3Without proper telemetry and performance metrics you will get to do this in a few more months again
Re: Saving months of compute time with a single Grafana query
#4I do not understand how cloud proponents talk about the he costs of self hosting but then get into situations like this.
Spending serious engineering time to wrangle with the complexities of cloud orchestration is not something that should be taken lightly.
Cloud services should be required to have a black-box Surgeon’s General warning.
Re: Saving months of compute time with a single Grafana query
#5Sadly Grafana (cloud) comes at a cost too. Anyone struggles with this horrible active metrics based pricing too? Not only Grafana Cloud but others do it like that too.
We moved shitloads to self hosted Thanos. While this comes with its own drawbacks obv, I think it was worth it.
Re: Saving months of compute time with a single Grafana query
#6I really enjoyed this read!
One thing that wasn't clear to me, is that if running NPM to install dependencies on pod startup is slow, why not pre build an image with dependencies already installed, and deploy that instead?
Re: Saving months of compute time with a single Grafana query
#7I do not understand how cloud proponents talk about the he costs of self hosting but then get into situations like this. Spending serious engineering time to wrangle with the complexities of cloud orchestration is not something that should be taken lightly. Cloud services should be required to have a black-box Surgeon’s General warning.
'accept vendor lock in, it'll save you the cost of engineers'
Routinely: oops, our API usage slipped and we mistakenly paid more than the staff to avoid this would cost
Keep fucking up, tech industry. My job role depends on it (SRE)
Re: Saving months of compute time with a single Grafana query
#8I'm really surprised that 300ms at startup would result in 25% fewer pods.... What % reduction in the total startup time is that?
Is it possible the prior measurement happened during a high traffic period and the post measurement happened in a low traffic period?
Re: Saving months of compute time with a single Grafana query
#9The 'one weird trick' could've been spotted in a graphical bundle analyser. But are they not caching npm packages somewhere, seems like an awful waste downloading from the npm registry over and over? I would think it would be parsing four different versions of the AWS sdk that was so slow.
Re: Saving months of compute time with a single Grafana query
#10I do not understand how cloud proponents talk about the he costs of self hosting but then get into situations like this. Spending serious engineering time to wrangle with the complexities of cloud orchestration is not something that should be taken lightly. Cloud services should be required to have a black-box Surgeon’s General warning.
'accept vendor lock in, it'll save you the cost of engineers' Routinely: oops, our API usage slipped and we mistakenly paid more than the staff to avoid this would cost Keep fucking up, tech industry. My job role depends on it (SRE)
The Tech Industry is well know for shunning the Price of Education, while rejecting the Cost of Ignorance.