4 ms·
Nice work: paper, (upcoming tools), and actual patches. Wish they'd look at disk I/O next, there are some problems there that are hard to describe other than a
by edwintorok 11y ago
Nice work: paper, (upcoming tools), and actual patches.
Wish they'd look at disk I/O next, there are some problems there that are hard to describe other than anecdotically: e.g. my system runs on a SSD and periodically copies data to a HDD with rsnapshot. When rsnapshot runs rm on the HDD things freeze for a moment, (even switching windows in X) although the only thing using the HDD is that rm ...
- EdiX 11y agoCould it be https://lkml.org/lkml/2016/3/30/424 https://lkml.org/lkml/2016/3/30/424 ?
- edwintorok 11y agothanks, I'll have to try that although in my case the writes and reads are done on entirely different devices (although maybe there is a shared queue somewhere?)
- melchebo 11y agoIt seems like everything with buffer/process queues needs a good polish to optimize for fairness. E.g. this is just being readied: http://blog.cerowrt.org/post/fq_codel_on_ath10k/ http://blog.cerowrt.org/post/fq_codel_on_ath10k/ (also note higher throughput as result)