I'm not sure about the client you're using, but I've noticed in the past the incorrect Thrift stack can make things run slow (like 40 times slower).
Check that the network stack wraps the socket in a Transport, preferably the TBufferedTransport. I'm guessing the client your're using is doing the right thing, just an suggestion. Aaron On Fri, 11 Jun 2010 18:49:46 -0700, "caribbean410" <caribbean...@gmail.com> wrote: > Thanks for the suggestion. For the test case, it is 1 key and 1 column. I > once changed 10 to 1, as I remember there is no much difference. > > > > I have 200k keys and each key is randomly generated. I will try the > optimized query next week. But maybe you still have to face the case that > each time a client just wants to query one key from db. > > > > From: Dop Sun [mailto:su...@dopsun.com] > Sent: Friday, June 11, 2010 6:05 PM > To: user@cassandra.apache.org > Subject: RE: read operation is slow > > > > And also, you are only select 1 key and 10 columns? > > > > criteria.keyList(Lists.newArrayList(userName)).columnRange(nameFirst, > nameFirst, 10); > > > > Then, if you have 200k keys, you have 200k Thrift calls. If this is the > case, you may need to optimize the way you do the query (to combine > multiple > keys into a single query), and to reduce the number of calls. > > > > From: Dop Sun [mailto:su...@dopsun.com] > Sent: Saturday, June 12, 2010 8:57 AM > To: user@cassandra.apache.org > Subject: RE: read operation is slow > > > > You mean after you "I remove some unnecessary column family and change the > size of rowcache and keycache, now the latency changes from 0.25ms to > 0.09ms. In essence 0.09ms*200k=18s.", it still takes 400 seconds to > returning? > > > > From: Caribbean410 [mailto:caribbean...@gmail.com] > Sent: Saturday, June 12, 2010 8:48 AM > To: user@cassandra.apache.org > Subject: Re: read operation is slow > > > > Hi, do you mean this one should not introduce much extra delay? To read a > record, I need select here, not sure where the extra delay comes from. > > On Fri, Jun 11, 2010 at 5:29 PM, Dop Sun <su...@dopsun.com> wrote: > > Jassandra is used here: > > > > Map<String, List<IColumn>> map = criteria.select(); > > > > The select here basically is a call to Thrift API: get_range_slices > > > > > > From: Caribbean410 [mailto:caribbean...@gmail.com] > Sent: Saturday, June 12, 2010 8:00 AM > > > To: user@cassandra.apache.org > Subject: Re: read operation is slow > > > > I remove some unnecessary column family and change the size of rowcache and > keycache, now the latency changes from 0.25ms to 0.09ms. In essence > 0.09ms*200k=18s. I don't know why it takes more than 400s total. Here is > the > client code and cfstats. There are not many operations here, why is the > extra time so large? > > > > long start = System.currentTimeMillis(); > for (int j = 0; j < 1; j++) { > for (int i = 0; i < numOfRecords; i++) { > int n = random.nextInt(numOfRecords); > ICriteria criteria = cf.createCriteria(); > userName = keySet[n]; > > criteria.keyList(Lists.newArrayList(userName)).columnRange(nameFirst, > nameFirst, 10); > Map<String, List<IColumn>> map = criteria.select(); > List<IColumn> list = map.get(userName); > // ByteArray bloc = list.get(0).getValue(); > // byte[] byteArrayloc = bloc.toByteArray(); > // loc = new String(byteArrayloc); > > // readBytes = readBytes + loc.length(); > readBytes = readBytes + blobSize; > } > } > > long finish=System.currentTimeMillis(); > > float totalTime=(finish-start)/1000; > > > Keyspace: Keyspace1 > Read Count: 600000 > Read Latency: 0.09053006666666667 ms. > Write Count: 200000 > Write Latency: 0.01504989 ms. > Pending Tasks: 0 > Column Family: Standard2 > SSTable count: 3 > Space used (live): 265990358 > Space used (total): 265990358 > Memtable Columns Count: 2615 > Memtable Data Size: 2667300 > Memtable Switch Count: 3 > Read Count: 600000 > Read Latency: 0.091 ms. > Write Count: 200000 > Write Latency: 0.015 ms. > Pending Tasks: 0 > Key cache capacity: 10000000 > Key cache size: 187465 > Key cache hit rate: 0.0 > Row cache capacity: 10000000 > Row cache size: 189990 > Row cache hit rate: 0.68335 > Compacted row minimum size: 0 > Compacted row maximum size: 0 > Compacted row mean size: 0 > > ---------------- > Keyspace: system > Read Count: 1 > Read Latency: 10.954 ms. > Write Count: 4 > Write Latency: 0.28075 ms. > Pending Tasks: 0 > Column Family: HintsColumnFamily > SSTable count: 0 > Space used (live): 0 > Space used (total): 0 > Memtable Columns Count: 0 > Memtable Data Size: 0 > Memtable Switch Count: 0 > Read Count: 0 > Read Latency: NaN ms. > Write Count: 0 > Write Latency: NaN ms. > Pending Tasks: 0 > Key cache capacity: 1 > Key cache size: 0 > Key cache hit rate: NaN > Row cache: disabled > Compacted row minimum size: 0 > Compacted row maximum size: 0 > Compacted row mean size: 0 > > Column Family: LocationInfo > SSTable count: 2 > Space used (live): 3232 > Space used (total): 3232 > Memtable Columns Count: 2 > Memtable Data Size: 46 > Memtable Switch Count: 1 > Read Count: 1 > Read Latency: 10.954 ms. > Write Count: 4 > Write Latency: 0.281 ms. > Pending Tasks: 0 > Key cache capacity: 1 > Key cache size: 1 > Key cache hit rate: 0.0 > Row cache: disabled > Compacted row minimum size: 0 > Compacted row maximum size: 0 > Compacted row mean size: 0 > > ---------------- > > On Fri, Jun 11, 2010 at 1:50 PM, Jonathan Ellis <jbel...@gmail.com> wrote: > > you need to look at cfstats to see what the latency is internal to > cassandra, vs what your client is introducing > > then you should probably read the comments in the configuration file > about caching > > > On Fri, Jun 11, 2010 at 9:38 AM, Caribbean410 <caribbean...@gmail.com> > wrote: >> >> Thanks Riyad. >> >> Right now I am just testing Cassandra on single node. The server and > client >> are running on the same machine. I tried the read test again on two >> machines, on one machine the cpu usage is around 30% most of the time and >> another is 90%. >> >> Pelops is one way to access Cassandra, there are also other java client > like >> hector and jassandra, will these java clients have significant different >> performance? >> >> Also I once tried to change the storage configure file, like change >> CommitLogDirectory and DataFileDirectory to different disks, change >> DiskAccessMode to mmap for a 64bit machine, and change ConcurrentReads > from >> 8 to 2. All of these do not change performance much. >> >> For other users who use different access client, like using php, c++, >> python, etc, if you have any experience in boosting the read performance, >> you are more than welcome to share with me. Thanks, >> >> On Fri, Jun 11, 2010 at 8:19 AM, Riyad Kalla <rka...@gmail.com> wrote: >>> >>> Caribbean410, >>> >>> This comes up on the Redis list alot as well -- what you are actually >>> measuring is the client sending a network connection to the Cas server > and >>> it replying -- so the performance numbers you are getting can easily be > 70% >>> network wait time and not necessarily hardcore read/write server >>> performance. >>> One way to see if this is the case, run your read test, then watch the > CPU >>> on the server for the Cassandra process and see if it's pegging the CPU > -- >>> if it's just sitting there banging between 0-10%, the you are spending > most >>> of your time waiting on network i/o (open/close sockets, etc.) >>> If you can parallelize your test to spawn say 5 threads that all do the >>> same thing, see if the performance for each thread increases linearly -- >>> which would indicate Cassandra is plenty fast in your setup, you just > need >>> to utilize more client threads over the network. >>> That new Java library, Pelops by Dominic >>> > (http://ria101.wordpress.com/2010/06/11/pelops-the-beautiful-cassandra-datab > ase-client-for-java/) >>> has a nice intrinsic node-balancing design that could be handy IF you >>> are >>> using multiple nodes. If you are just testing against 1 node, then spawn >>> multiple threads of your code above and see how each thread's >>> performance >>> scales. >>> -R >>> On Thu, Jun 10, 2010 at 2:39 PM, Caribbean410 <caribbean...@gmail.com> >>> wrote: >>>> >>>> Hello, >>>> >>>> I am testing the performance of cassandra. We write 200k records to >>>> database and each record is 1k size. Then we read these 200k records. >>>> It takes more than 400s to finish the read which is much slower than >>>> mysql (20s around). I read some discussion online and someone suggest >>>> to make multiple connections to make it faster. But I am not sure how >>>> to do it, do I need to change my storage setting file or just change >>>> the java client code? >>>> >>>> Here is my read code, >>>> >>>> Properties info = new Properties(); >>>> info.put(DriverManager.CONSISTENCY_LEVEL, >>>> ConsistencyLevel.ONE.toString()); >>>> >>>> IConnection connection = > DriverManager.getConnection( >>>> "thrift://localhost:9160", info); >>>> >>>> // 2. Get a KeySpace by name >>>> IKeySpace keySpace = >>>> connection.getKeySpace("Keyspace1"); >>>> >>>> // 3. Get a ColumnFamily by name >>>> IColumnFamily cf = >>>> keySpace.getColumnFamily("Standard2"); >>>> >>>> ByteArray nameFirst = ByteArray.ofASCII("first"); >>>> ICriteria criteria = cf.createCriteria(); >>>> long readBytes = 0; >>>> long start = System.currentTimeMillis(); >>>> for (int i = 0; i < numOfRecords; i++) { >>>> int n = random.nextInt(numOfRecords); >>>> userName = keySet[n]; >>>> >>>> criteria.keyList(Lists.newArrayList(userName)).columnRange(nameFirst, >>>> nameFirst, 10); >>>> Map<String, List<IColumn>> map = >>>> criteria.select(); >>>> List<IColumn> list = >>>> map.get(userName); >>>> ByteArray bloc = >>>> list.get(0).getValue(); >>>> byte[] byteArrayloc = >>>> bloc.toByteArray(); >>>> loc = new String(byteArrayloc); >>>> // System.out.println(userName+" >>>> "+loc); >>>> readBytes = readBytes + >>>> loc.length(); >>>> } >>>> >>>> long finish=System.currentTimeMillis(); >>>> >>>> I once commented these lines >>>> >>>> ByteArray bloc = >>>> list.get(0).getValue(); >>>> byte[] byteArrayloc = >>>> bloc.toByteArray(); >>>> loc = new String(byteArrayloc); >>>> // System.out.println(userName+" >>>> "+loc); >>>> readBytes = readBytes + >>>> loc.length(); >>>> >>>> And the performance doesn't improve much. >>>> >>>> Any suggestion is welcome. Thanks, >> >> > > -- > Jonathan Ellis > Project Chair, Apache Cassandra > co-founder of Riptano, the source for professional Cassandra support > http://riptano.com
