I found the problem; it was my fault. I made an accidental change to my
cassandra.yaml file sometime between restarts and ended up pointing the
node data directory to a different disk. Check your paths!
Todd
On 2/10/2012 10:22 AM, Todd Fast wrote:
My single-node cluster was working fine yesterday. I ctrl+c'd it last
night, as I typically do, and restarted it this morning.
Now, inexplicably, it doesn't know anything about my keyspace. The SS
table files are in the same directory as always and seem to be the
expected size. I can't seem to do anything with nodetool, since the
keyspace isn't known.
1. How can I recover the node?
2. What the heck happened that caused this?
Here is the console log (I'm on Windows 7):
Starting Cassandra Server
INFO 09:33:41,803 Logging initialized
INFO 09:33:41,809 JVM vendor/version: Java HotSpot(TM) Client
VM/1.6.0_17
INFO 09:33:41,809 Heap size: 1065484288/1065484288
INFO 09:33:41,810 Classpath: ...
INFO 09:33:41,815 JNA not found. Native methods will be disabled.
INFO 09:33:41,826 Loading settings from
file:/D:/Java/apache-cassandra-1.0.7/conf/cassandra.yaml
INFO 09:33:41,930 DiskAccessMode 'auto' determined to be standard,
indexAccessMode is standard
INFO 09:33:41,939 Global memtable threshold is enabled at 338MB
INFO 09:33:42,232 Opening
\Data\cassandra\node1\system\LocationInfo-hc-1 (234 bytes)
INFO 09:33:42,232 Opening
\Data\cassandra\node1\system\LocationInfo-hc-2 (163 bytes)
INFO 09:33:42,287 Couldn't detect any schema definitions in local
storage.
INFO 09:33:42,288 Found table data in data directories. Consider
using the CLI to define your schema.
INFO 09:33:42,307 Creating new commitlog segment
/Data/cassandra/node1/commitlog\CommitLog-1328895222306.log
INFO 09:33:42,316 Replaying
\Data\cassandra\node1\commitlog\CommitLog-1328894913967.log
INFO 09:33:42,356 Finished reading
\Data\cassandra\node1\commitlog\CommitLog-1328894913967.log
INFO 09:33:42,362 Enqueuing flush of
Memtable-Versions@22744620(83/103 serialized/live bytes, 3 ops)
INFO 09:33:42,364 Writing Memtable-Versions@22744620(83/103
serialized/live bytes, 3 ops)
INFO 09:33:42,399 Completed flushing
\Data\cassandra\node1\system\Versions-hc-1-Data.db (247 bytes)
INFO 09:33:42,410 Log replay complete, 3 replayed mutations
INFO 09:33:42,415 Cassandra version: 1.0.7
INFO 09:33:42,416 Thrift API version: 19.20.0
INFO 09:33:42,416 Loading persisted ring state
INFO 09:33:42,420 Starting up server gossip
INFO 09:33:42,429 Enqueuing flush of
Memtable-LocationInfo@18721294(29/36 serialized/live bytes, 1 ops)
INFO 09:33:42,430 Writing Memtable-LocationInfo@18721294(29/36
serialized/live bytes, 1 ops)
INFO 09:33:42,450 Completed flushing
\Data\cassandra\node1\system\LocationInfo-hc-3-Data.db (80 bytes)
INFO 09:33:42,459 Starting Messaging Service on port 7000
INFO 09:33:42,469 Using saved token
133677729504783243750441433892785690257
INFO 09:33:42,471 Enqueuing flush of
Memtable-LocationInfo@15427560(53/66 serialized/live bytes, 2 ops)
INFO 09:33:42,471 Writing Memtable-LocationInfo@15427560(53/66
serialized/live bytes, 2 ops)
INFO 09:33:42,490 Completed flushing
\Data\cassandra\node1\system\LocationInfo-hc-4-Data.db (163 bytes)
INFO 09:33:42,494 Node localhost/127.0.0.1 state jump to normal
INFO 09:33:42,511 Bootstrap/Replace/Move completed! Now serving reads.
INFO 09:33:42,512 Will not load MX4J, mx4j-tools.jar is not in the
classpath
INFO 09:33:42,523 Compacting
[SSTableReader(path='\Data\cassandra\node1\system\LocationInfo-hc-4-Data.db'),
SSTableReader(path='\Data\cassandra\node1\system\LocationInfo-hc-2-Data.db'),
SSTableReader(path='\Data\cassandra\node1\system\Loca
tionInfo-hc-3-Data.db'),
SSTableReader(path='\Data\cassandra\node1\system\LocationInfo-hc-1-Data.db')]
INFO 09:33:42,567 Binding thrift service to localhost/127.0.0.1:9160
INFO 09:33:42,571 Using TFastFramedTransport with a max frame size of
15728640 bytes.
INFO 09:33:42,576 Using synchronous/threadpool thrift server on
localhost/127.0.0.1 : 9160
INFO 09:33:42,577 Listening for thrift clients...
Todd
