Hello! Have you tried setting queryParallelism on this cache to some high value (such as your number of CPU cores)?
Note that this query will likely cause lifting all data in cache to heap at once, which is probably not what you expect. My recommendation is also to try splitting this operation into smaller ones, see if they finish faster. Regards, -- Ilya Kasnacheev вт, 9 окт. 2018 г. в 15:47, Justin Ji <binarytre...@foxmail.com>: > Hi - > > I have a cache with more than 1 million records, when I update the whole > records in it, it spends a long time(almost 6 minutes). > > Here is the SQL: > "update " + IgniteTableKey.T_DEVICE_ONLINE_STATUS.getCode()+ " set > isOnline=0, mqttTime=" + System.currentTimeMillis() / 1000 + " where 1=1"; > > And the Cache configuration is : > > CacheConfiguration<K, V> cacheCfg = new CacheConfiguration<>(); > cacheCfg.setName(cacheName); > cacheCfg.setCacheMode(CacheMode.PARTITIONED); > cacheCfg.setBackups(1); > cacheCfg.setAtomicityMode(CacheAtomicityMode.TRANSACTIONAL); > > > cacheCfg.setCacheStoreFactory(FactoryBuilder.factoryOf(DeviceStatusCacheStore.class)); > cacheCfg.setWriteThrough(true); > cacheCfg.setWriteBehindEnabled(true); > cacheCfg.setReadThrough(true); > cacheCfg.setWriteBehindFlushThreadCount(4); > cacheCfg.setWriteBehindFlushFrequency(15 * 1000); > cacheCfg.setWriteBehindFlushSize(409600); > cacheCfg.setWriteBehindBatchSize(1024); > cacheCfg.setStoreKeepBinary(true); > cfg.setCacheConfiguration(cacheCfg); > > So I want to know is there a way that can speed up the execution？ > > > > -- > Sent from: http://apache-ignite-users.70518.x6.nabble.com/ >
