Home | About | Sematext search-lucene.com search-hadoop.com
 Search Hadoop and all its subprojects:

Switch to Threaded View
Accumulo, mail # user - Thrift error while configuring new cloud


Copy link to this message
-
Re: Thrift error while configuring new cloud
Keith Turner 2013-01-17, 14:59
I think the default stack size is 1M, or it used to be.  Anyway 128M
is probably too large.   I think the your best option is to just
remove the -Xss option.

On Wed, Jan 16, 2013 at 4:43 PM, Hider, Sandy <[EMAIL PROTECTED]> wrote:
> Thanks Eric and John for the replies.  In case anyone else runs into this
> problem.
>
>
>
> On the slave nodes, in the accumulo log dir, in file
> tserver_<computer_name>.out it stated.
>
>
>
> The stack size specified is too small, Specify at least 160k
>
>
>
> I had accidentally put a k instead of an m on the tserver line in
> accumulo/conf/accumulo-env.sh
>
>
>
> test -z "$ACCUMULO_TSERVER_OPTS" && export ACCUMULO_TSERVER_OPTS="${POLICY}
> -Xmx500m -Xms128m -Xss128k"
>
> instead of
>
> test -z "$ACCUMULO_TSERVER_OPTS" && export ACCUMULO_TSERVER_OPTS="${POLICY}
> -Xmx500m -Xms128m -Xss128m"
>
>
>
> Thanks again,
>
>
>
> Sandy
>
>
>
>
>
>
>
> From: Eric Newton [mailto:[EMAIL PROTECTED]]
> Sent: Monday, January 07, 2013 4:57 PM
> To: [EMAIL PROTECTED]; vines
>
>
> Subject: Re: Thrift error while configuring new cloud
>
>
>
> Make sure the $ACCUMULO_LOG_DIR exists on all the nodes, too.
>
>
>
> -Eric
>
>
>
>
>
> On Mon, Jan 7, 2013 at 4:48 PM, John Vines <[EMAIL PROTECTED]> wrote:
>
> If you actually visit the monitor page, on port 50095, it does log
> aggregation so you don't have to dig up logs.
>
> It looks like your tserver and logger are not running. This is why things
> are not working for you. There may be log info on the monitor page. But you
> may have to go to your logs directory and find the logs for them. There may
> only be a .out and .err file for them, which should describe the error.
>
>
>
> On Mon, Jan 7, 2013 at 4:34 PM, Hider, Sandy <[EMAIL PROTECTED]> wrote:
>
> In monitor_<servername>.log
>
>
>
> 04 16:56:46,085 [server.Accumulo] INFO :  tserver.bloom.load.concurrent.max
> = 4
>
> 04 16:56:46,085 [server.Accumulo] INFO :  tserver.bulk.assign.threads = 1
>
> 04 16:56:46,085 [server.Accumulo] INFO :  tserver.bulk.process.threads = 1
>
> 04 16:56:46,085 [server.Accumulo] INFO :  tserver.bulk.retry.max = 3
>
> 04 16:56:46,085 [server.Accumulo] INFO :  tserver.cache.data.size = 50M
>
> 04 16:56:46,086 [server.Accumulo] INFO :  tserver.cache.index.size = 512M
>
> 04 16:56:46,086 [server.Accumulo] INFO :  tserver.client.timeout = 3s
>
> 04 16:56:46,086 [server.Accumulo] INFO :
> tserver.compaction.major.concurrent.max = 3
>
> 04 16:56:46,088 [server.Accumulo] INFO :  tserver.compaction.major.delay > 30s
>
> 04 16:56:46,088 [server.Accumulo] INFO :
> tserver.compaction.major.thread.files.open.max = 10
>
> 04 16:56:46,089 [server.Accumulo] INFO :
> tserver.compaction.minor.concurrent.max = 4
>
> 04 16:56:46,089 [server.Accumulo] INFO :  tserver.default.blocksize = 1M
>
> 04 16:56:46,089 [server.Accumulo] INFO :  tserver.dir.memdump = /tmp
>
> 04 16:56:46,089 [server.Accumulo] INFO :  tserver.files.open.idle = 1m
>
> 04 16:56:46,089 [server.Accumulo] INFO :  tserver.hold.time.max = 5m
>
> 04 16:56:46,089 [server.Accumulo] INFO :  tserver.logger.count = 2
>
> 04 16:56:46,089 [server.Accumulo] INFO :  tserver.logger.strategy > org.apache.accumulo.server.tabletserver.log.RoundRobinLoggerStrategy
>
> 04 16:56:46,090 [server.Accumulo] INFO :  tserver.logger.timeout = 30s
>
> 04 16:56:46,090 [server.Accumulo] INFO :  tserver.memory.lock = false
>
> 04 16:56:46,090 [server.Accumulo] INFO :  tserver.memory.manager > org.apache.accumulo.server.tabletserver.LargestFirstMemoryManager
>
> 04 16:56:46,090 [server.Accumulo] INFO :  tserver.memory.maps.max = 1G
>
> 04 16:56:46,090 [server.Accumulo] INFO :  tserver.memory.maps.native.enabled
> = true
>
> 04 16:56:46,090 [server.Accumulo] INFO :
> tserver.metadata.readahead.concurrent.max = 8
>
> 04 16:56:46,090 [server.Accumulo] INFO :  tserver.migrations.concurrent.max
> = 1
>
> 04 16:56:46,091 [server.Accumulo] INFO :  tserver.monitor.fs = true
>
> 04 16:56:46,091 [server.Accumulo] INFO :  tserver.mutation.queue.max = 256K
>
> 04 16:56:46,091 [server.Accumulo] INFO :  tserver.port.client = 9997