чтобы быть увереным что читаем с мастера! и данные не проебутся если этот мастер сдохнет!
В ТРИ СУКА ЧЕТЫРЕ
scontent-b-ams.xx.fbcdn.net
на фотографии справа снизу чувак рассказывает про монгодб
теперь посмотрите на лица слушателей
на фотографии справа снизу чувак рассказывает про монгодб
теперь посмотрите на лица слушателей
на 285 машинах с ссд
по 3к рпс пер ноду. хайлоад
MongoDB: create the right indexes for your queries; On Linux, choose ext4 or xfs; Since MongoDB is constantly accessing its files, you can get significant performance by telling Linux not to track files' access times; calculate working set size correctly; use SSD when storing data large than RAM; shard.
так смешно стало
▪ Five Things About Scaling так смешно стало
drive.google.com
тлдр: 15кк Документов, четыре ноды, 10г интернет
лучший результат у коучбейза, кстати не путайте с коучдб, в 22к рпс!
ГРАУНД БРЕЙКИНГ
consistency for compare-and-set operations. The implementation is based on naive Paxos–requiring four round trips for each write.
The first thing you’ll notice if you try to test the new transaction system is that the Java driver doesn’t support it. It’ll throw some weird exceptions like “unknown consistency level SERIAL”, because it doesn’t support the v2 native Cassandra protocol yet. So you’ll need to use the Python Thrift client, or, in my case, get a patched client from DataStax.
The second thing you’ll notice is deadlocks. In my Jepsen tests, the cluster would go unresponsive after the first 10 or so transactions–and it would never recover. Any further attempts to modify a cell via transaction would spin endlessly in failed transactions, until I manually truncated the system.paxos table.
In Cassandra 2.0.0, Lightweight Transactions offer linearizable The first thing you’ll notice if you try to test the new transaction system is that the Java driver doesn’t support it. It’ll throw some weird exceptions like “unknown consistency level SERIAL”, because it doesn’t support the v2 native Cassandra protocol yet. So you’ll need to use the Python Thrift client, or, in my case, get a patched client from DataStax.
The second thing you’ll notice is deadlocks. In my Jepsen tests, the cluster would go unresponsive after the first 10 or so transactions–and it would never recover. Any further attempts to modify a cell via transaction would spin endlessly in failed transactions, until I manually truncated the system.paxos table.
и заменили его пер датабейз врайт локом
ПРОРЫВ ИННОВАЦИИ