Thrift - CQL
Hi all, We have been using Cassandra for more than 3 years and now we have a cluster in production still running on 1.1.x contains dynamic-columned column-families - with hector as client. We are trying to update to the latest 1.2.x and considering to use datastax client in order to utilise some of its round robin / failover goodness. We bumped on to a few walls however when converting our thrift based client code to CQL. We read through the docs + datastax dev blog entires like: this and this. However they are mostly focus on reading from an existing dynamic cf, run some alter table statements, and reading it again. Very little about how to insert / update. So there comes my questions: - Is there any way to do insert / update at all on a good old wide cf using CQL? Based on what we read back out, we have tried: INSERT INTO cf_name(key, column1, value) VALUES (‘key1’, ‘columnName1’,’columnValue2’) But we ended up with “Unknown identifier column1” - About read - One of our cf is defined with a secondary index. So the schema looks something like: create column family cf_with_index with column_type = 'Standard' and comparator = 'UTF8Type' and default_validation_class = 'UTF8Type' and key_validation_class = 'UTF8Type' and column_metadata = [ {column_name : 'indexed_column', validation_class : UTF8Type, index_name : 'column_idx', index_type : 0}]; When reading from cli, we will see all columns, data as you expected: -- --- RowKey: rowkey1 = (name=c1, value=v1, timestamp=xxx, ttl=604800) = (name=c2, value=v2, timestamp=xxx, ttl=604800) = (name=c3, value=v3, timestamp=xxx, ttl=604800) = (name=indexed_column, value=value1, timestamp=xxx, ttl=604800) --- However when we Query via CQL, we only get the indexed column: SELECT * FROM cf_with_index WHERE key = ‘rowkey1’; key | indexed_column ---+ rowkey1 | value1 Any way to get the rest? - Obtaining TTL and writetime on these wide rows - we tried: SELECT key, column1, value, writetime(value), ttl(value) FROM cf LIMIT 1; It works, but a bit clumsy. Is there a better way? - We can live with thrift. Is there any way / plan to let us to execute thrift with datastax driver? Hector seems not active anymore. Many thanks in advanced, A
Re: Commit log periodic sync?
Thanks again Aaron. I think case I would not expect to see data lose. If you are still in a test scenario can you try to reproduce the problem ? If possible can you reproduce it with a single node ? We will try that later this week. We did the same exercise this week, this time we did a flush and snapshot before the DR actually happened - as an attempt to identify if the commit logs fsync was the problem. We can clearly see stables were created for the flush command. And those sstables were loaded in when the nodes started up again after the DR exercise. At this point we believed all nodes had all the data, so we let them serving client requests while we run repair on the nodes. Data created before the last flush was still missing, according to the client that talked to DC1 (the disaster DC). We had a look at the log of one of the DC1 nodes. The suspicious thing was that latest sstable was being compacted during streaming sessions of the repair. But no error was reported. Here comes my questions: - if during the streaming session, the sstable that was about to stream out but was being compacted, would we see error in the log? - could this lead to data not found? - is it safe to let a node serving read/write requests while repair is running? Many thanks again. - A aaron morton aa...@thelastpickle.com 於 27 Aug 2012 09:08 寫道: Brutally. kill -9. that's fine. I was thinking about reboot -f -n We are wondering if the fsync of the commit log was working. I would say yes only because there other reported problems. I think case I would not expect to see data lose. If you are still in a test scenario can you try to reproduce the problem ? If possible can you reproduce it with a single node ? Cheers - Aaron Morton Freelance Developer @aaronmorton http://www.thelastpickle.com On 25/08/2012, at 11:00 AM, rubbish me rubbish...@googlemail.com wrote: Thanks, Aaron, for your reply - please see the inline. On 24 Aug 2012, at 11:04, aaron morton wrote: - we are running on production linux VMs (not ideal but this is out of our hands) Is the VM doing anything wacky with the IO ? Could be. But I thought we would ask here first. This is a bit difficult to prove cos we dont have the control over these VMs. As part of a DR exercise, we killed all 6 nodes in DC1, Nice disaster. Out of interest, what was the shutdown process ? Brutally. kill -9. We noticed that data that was written an hour before the exercise, around the last memtables being flushed,was not found in DC1. To confirm, data was written to DC 1 at CL LOCAL_QUORUM before the DR exercise. Was the missing data written before or after the memtable flush ? I'm trying to understand if the data should have been in the commit log or the memtables. Missing data was those written after the last flush. These data was retrievable before the DR exercise. Can you provide some more info on how you are detecting it is not found in DC 1? We tried hector, consistencylevel=local quorum. We had missing column or the whole row. We tried cassandra-cli on DC1 nodes, same. However once we run the same query on DC2, C* must have then done a read-repair. That particular piece of result data would appear in DC1 again. If we understand correctly, commit logs are being written first and then to disk every 10s. Writes are put into a bounded queue and processed as fast as the IO can keep up. Every 10s a sync messages is added to the queue. Not that the commit log segment may rotate at any time which requires a sync. A loss of data across all nodes in a DC seems odd. If you can provide some more information we may be able to help. We are wondering if the fsync of the commit log was working. But we saw no errors / warning in logs. Wondering if there is way to verify Cheers - Aaron Morton Freelance Developer @aaronmorton http://www.thelastpickle.com On 24/08/2012, at 6:01 AM, rubbish me rubbish...@googlemail.com wrote: Hi all First off, let's introduce the setup. - 6 x C* 1.1.2 in active DC (DC1), another 6 in another (DC2) - keyspace's RF=3 in each DC - Hector as client. - client talks only to DC1 unless DC1 can't serve the request. In which case talks only to DC2 - commit log was periodically sync with the default setting of 10s. - consistency policy = LOCAL QUORUM for both read and write. - we are running on production linux VMs (not ideal but this is out of our hands) - As part of a DR exercise, we killed all 6 nodes in DC1, hector starts talking to DC2, all the data was still there, everything continued to work perfectly. Then we brought all nodes, one by one, in DC1 up. We saw a message saying all the commit logs were replayed. No errors reported. We didn't run repair at this time. We noticed that data
commit log to disk with periodic mode
Hi all First off, please let me introduce the setup. - a balance ring of 6 x C* 1.1.2 in active DC (DC1), 6 in another (DC2); - keyspace's RF=3 in each DC; - client talks only to DC1 unless DC1 can't serve the request, in which case talks only to DC2; - commit log is being sync periodically with the default setting of 10s. - consistency policy = LOCAL QUORUM for both read and write. - we are running on production linux VMs (not ideal but this is out of our hands) - As part of a DR exercise, we brutally killed all 6 nodes in DC1, client started talking to DC2. All data survived, everything continued to work perfectly. Then we brought all nodes in DC1 up, one by one We saw each with message saying commit logs were all replayed. No errors reported. We didn't run repair at this time. However, DC1 lost data that was written an hour before the DR exercise. It seemed everything after the last memtable-flush was gone. If we understand correctly, commit logs are being written first and then sync to disk every 10s. At worst we would have lost the last 10s of data. But it seemed as if the periodic sync didnt happen. What could be the cause of this behaviour? With the blessing of C* we could recovered all these data from DC2. But we would like to understand the possible cause. Many thanks in advanced. - A
Re: BulkLoading sstables from v1.0.3 to v1.1.1
Thanks Ivo. We are quite close to releasing so we'd hope to understand what causing the error and may try to avoid it where possible. As said, it seems to work ok the first time round. The problem you referring in the last mail, was it restricted to bulk loading or otherwise? Thanks -A Ivo Meißner i...@overtronic.com 於 10 Jul 2012 07:20 寫道: Hi, there are some problems in version 1.1.1 with secondary indexes and key caches that are fixed in 1.1.2. I would try to upgrade to 1.1.2 and see if the error still occurs. Ivo Hi As part of a continuous development of a system migration, we have a test build to take a snapshot of a keyspace from cassandra v 1.0.3 and bulk load it to a cluster of 1.1.1 using the sstableloader.sh. Not sure if relevant, but one of the cf contains a secondary index. The build basically does: Drop the destination keyspace if exist Add the destination keyspace, wait for schema to agree run sstableLoader Do some validation of the streamed data Keyspace / column families schema are basically the same, apart from in the one of v1.1.1, we had compression and key cache switched on. On a clean cluster, (empty data, commit log, saved-cache dirs) the sstables loaded beautifully. But subsequent build failed with -- [21:02:02][exec] progress: [snip ip_addresses]... [total: 0 - 0MB/s (avg: 0MB/s)]ERROR 21:02:02,811 Error in ThreadPoolExecutorjava.lang.RuntimeException: java.net.SocketException: Connection reset