You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@zookeeper.apache.org by Joe Meszaros <jo...@impresstv.com> on 2015/03/26 12:06:16 UTC
Zookeeper 3.5 stable
Hi,
I've seen that you plan to make 3.5.1 RC soon, but it is not clear for me
when do you plan to publish the 3.5.X stable version. I did not found
usable information in the archived messages, nor on the internet.
I am curious because we shipped a production system with Zookeeper and we
will need the dynamic cluster reconfiguration.
Many thanks,
Joe
Re: ZooKeeper
Posted by Chris Barlock <ba...@us.ibm.com>.
Thanks, Vikrant. We will look into this.
Chris
From: Vikrant Singh <vi...@yahoo-inc.com.INVALID>
To: "user@zookeeper.apache.org" <us...@zookeeper.apache.org>
Date: 04/16/2015 03:08 PM
Subject: Re: ZooKeeper
Not sure, but this may be problem..<mem type="nursery" free="0"
total="120782848" percent="0" />
You may try to adjust memory assigned to JVM. I think -Xns is the option
you may like to use.
On Thursday, 16 April 2015 10:52 AM, Chris Barlock
<ba...@us.ibm.com> wrote:
We have a system which is running a single instance of ZK. It is
currently using 100% of a CPU core. The zookeeper-server.err file is
logging one of these entries every two seconds:
<exclusive-start id="699575" timestamp="2015-04-16T13:43:10.361"
intervalms="2179.021">
<response-info timems="0.011" idlems="0.011" threads="0"
lastid="000000002152F500"
lastname="NIOServerCxn.Factory:0.0.0.0/0.0.0.0:2181" />
</exclusive-start>
<af-start id="699576" totalBytesRequested="16392"
timestamp="2015-04-16T13:43:10.361" intervalms="2179.019" />
<cycle-start id="699577" type="scavenge" contextid="0"
timestamp="2015-04-16T13:43:10.361" intervalms="2179.017" />
<gc-start id="699578" type="scavenge" contextid="699577"
timestamp="2015-04-16T13:43:10.361">
<mem-info id="699579" free="325374712" total="523370496" percent="62">
<mem type="nursery" free="0" total="120782848" percent="0" />
<mem type="tenure" free="325374712" total="402587648" percent="80">
<mem type="soa" free="324569848" total="401782784" percent="80" />
<mem type="loa" free="804864" total="804864" percent="100" />
</mem>
<remembered-set count="1964" />
</mem-info>
</gc-start>
<allocation-stats totalBytes="120125544" >
<allocated-bytes non-tlh="16392" tlh="120109152" />
<largest-consumer threadName="NIOServerCxn.Factory:0.0.0.0/0.0.0.0:2181"
threadId="000000002152F500" bytes="118617128" />
</allocation-stats>
<gc-op id="699580" type="scavenge" timems="1.907" contextid="699577"
timestamp="2015-04-16T13:43:10.363">
<scavenger-info tenureage="14" tiltratio="89" />
<memory-copied type="nursery" objects="1099" bytes="93088"
bytesdiscarded="300896" />
<memory-copied type="tenure" objects="12" bytes="8464"
bytesdiscarded="122608" />
</gc-op>
<gc-end id="699581" type="scavenge" contextid="699577" durationms="1.977"
timestamp="2015-04-16T13:43:10.363">
<mem-info id="699582" free="445632504" total="523370496" percent="85">
<mem type="nursery" free="120388864" total="120782848" percent="99" />
<mem type="tenure" free="325243640" total="402587648" percent="80">
<mem type="soa" free="324438776" total="401782784" percent="80" />
<mem type="loa" free="804864" total="804864" percent="100" />
</mem>
<remembered-set count="107" />
</mem-info>
</gc-end>
<cycle-end id="699583" type="scavenge" contextid="699577"
timestamp="2015-04-16T13:43:10.363" />
<allocation-satisfied id="699584" threadId="000000002152F500"
bytesRequested="16392" />
<af-end id="699585" timestamp="2015-04-16T13:43:10.363" />
<exclusive-end id="699586" timestamp="2015-04-16T13:43:10.363"
durationms="2.177" />
I can connec to the server with zkCli.sh, but a simple "ls /" fails with a
ConnectionLossException. I have no idea how to go about diagnosing the
problem, so any help would be appreciated.
Thanks!
Chris
Re: ZooKeeper
Posted by Vikrant Singh <vi...@yahoo-inc.com.INVALID>.
Not sure, but this may be problem..<mem type="nursery" free="0" total="120782848" percent="0" />
You may try to adjust memory assigned to JVM. I think -Xns is the option you may like to use.
On Thursday, 16 April 2015 10:52 AM, Chris Barlock <ba...@us.ibm.com> wrote:
We have a system which is running a single instance of ZK. It is
currently using 100% of a CPU core. The zookeeper-server.err file is
logging one of these entries every two seconds:
<exclusive-start id="699575" timestamp="2015-04-16T13:43:10.361"
intervalms="2179.021">
<response-info timems="0.011" idlems="0.011" threads="0"
lastid="000000002152F500"
lastname="NIOServerCxn.Factory:0.0.0.0/0.0.0.0:2181" />
</exclusive-start>
<af-start id="699576" totalBytesRequested="16392"
timestamp="2015-04-16T13:43:10.361" intervalms="2179.019" />
<cycle-start id="699577" type="scavenge" contextid="0"
timestamp="2015-04-16T13:43:10.361" intervalms="2179.017" />
<gc-start id="699578" type="scavenge" contextid="699577"
timestamp="2015-04-16T13:43:10.361">
<mem-info id="699579" free="325374712" total="523370496" percent="62">
<mem type="nursery" free="0" total="120782848" percent="0" />
<mem type="tenure" free="325374712" total="402587648" percent="80">
<mem type="soa" free="324569848" total="401782784" percent="80" />
<mem type="loa" free="804864" total="804864" percent="100" />
</mem>
<remembered-set count="1964" />
</mem-info>
</gc-start>
<allocation-stats totalBytes="120125544" >
<allocated-bytes non-tlh="16392" tlh="120109152" />
<largest-consumer threadName="NIOServerCxn.Factory:0.0.0.0/0.0.0.0:2181"
threadId="000000002152F500" bytes="118617128" />
</allocation-stats>
<gc-op id="699580" type="scavenge" timems="1.907" contextid="699577"
timestamp="2015-04-16T13:43:10.363">
<scavenger-info tenureage="14" tiltratio="89" />
<memory-copied type="nursery" objects="1099" bytes="93088"
bytesdiscarded="300896" />
<memory-copied type="tenure" objects="12" bytes="8464"
bytesdiscarded="122608" />
</gc-op>
<gc-end id="699581" type="scavenge" contextid="699577" durationms="1.977"
timestamp="2015-04-16T13:43:10.363">
<mem-info id="699582" free="445632504" total="523370496" percent="85">
<mem type="nursery" free="120388864" total="120782848" percent="99" />
<mem type="tenure" free="325243640" total="402587648" percent="80">
<mem type="soa" free="324438776" total="401782784" percent="80" />
<mem type="loa" free="804864" total="804864" percent="100" />
</mem>
<remembered-set count="107" />
</mem-info>
</gc-end>
<cycle-end id="699583" type="scavenge" contextid="699577"
timestamp="2015-04-16T13:43:10.363" />
<allocation-satisfied id="699584" threadId="000000002152F500"
bytesRequested="16392" />
<af-end id="699585" timestamp="2015-04-16T13:43:10.363" />
<exclusive-end id="699586" timestamp="2015-04-16T13:43:10.363"
durationms="2.177" />
I can connec to the server with zkCli.sh, but a simple "ls /" fails with a
ConnectionLossException. I have no idea how to go about diagnosing the
problem, so any help would be appreciated.
Thanks!
Chris
ZooKeeper
Posted by Chris Barlock <ba...@us.ibm.com>.
We have a system which is running a single instance of ZK. It is
currently using 100% of a CPU core. The zookeeper-server.err file is
logging one of these entries every two seconds:
<exclusive-start id="699575" timestamp="2015-04-16T13:43:10.361"
intervalms="2179.021">
<response-info timems="0.011" idlems="0.011" threads="0"
lastid="000000002152F500"
lastname="NIOServerCxn.Factory:0.0.0.0/0.0.0.0:2181" />
</exclusive-start>
<af-start id="699576" totalBytesRequested="16392"
timestamp="2015-04-16T13:43:10.361" intervalms="2179.019" />
<cycle-start id="699577" type="scavenge" contextid="0"
timestamp="2015-04-16T13:43:10.361" intervalms="2179.017" />
<gc-start id="699578" type="scavenge" contextid="699577"
timestamp="2015-04-16T13:43:10.361">
<mem-info id="699579" free="325374712" total="523370496" percent="62">
<mem type="nursery" free="0" total="120782848" percent="0" />
<mem type="tenure" free="325374712" total="402587648" percent="80">
<mem type="soa" free="324569848" total="401782784" percent="80" />
<mem type="loa" free="804864" total="804864" percent="100" />
</mem>
<remembered-set count="1964" />
</mem-info>
</gc-start>
<allocation-stats totalBytes="120125544" >
<allocated-bytes non-tlh="16392" tlh="120109152" />
<largest-consumer threadName="NIOServerCxn.Factory:0.0.0.0/0.0.0.0:2181"
threadId="000000002152F500" bytes="118617128" />
</allocation-stats>
<gc-op id="699580" type="scavenge" timems="1.907" contextid="699577"
timestamp="2015-04-16T13:43:10.363">
<scavenger-info tenureage="14" tiltratio="89" />
<memory-copied type="nursery" objects="1099" bytes="93088"
bytesdiscarded="300896" />
<memory-copied type="tenure" objects="12" bytes="8464"
bytesdiscarded="122608" />
</gc-op>
<gc-end id="699581" type="scavenge" contextid="699577" durationms="1.977"
timestamp="2015-04-16T13:43:10.363">
<mem-info id="699582" free="445632504" total="523370496" percent="85">
<mem type="nursery" free="120388864" total="120782848" percent="99" />
<mem type="tenure" free="325243640" total="402587648" percent="80">
<mem type="soa" free="324438776" total="401782784" percent="80" />
<mem type="loa" free="804864" total="804864" percent="100" />
</mem>
<remembered-set count="107" />
</mem-info>
</gc-end>
<cycle-end id="699583" type="scavenge" contextid="699577"
timestamp="2015-04-16T13:43:10.363" />
<allocation-satisfied id="699584" threadId="000000002152F500"
bytesRequested="16392" />
<af-end id="699585" timestamp="2015-04-16T13:43:10.363" />
<exclusive-end id="699586" timestamp="2015-04-16T13:43:10.363"
durationms="2.177" />
I can connec to the server with zkCli.sh, but a simple "ls /" fails with a
ConnectionLossException. I have no idea how to go about diagnosing the
problem, so any help would be appreciated.
Thanks!
Chris
Re: Zookeeper 3.5 stable
Posted by Patrick Hunt <ph...@apache.org>.
Certainly not a month. We're working on 3.5.1-alpha right now.
Patrick
On Tue, Mar 31, 2015 at 1:53 AM, József Mészáros
<jo...@impresstv.com> wrote:
> Hi Patrick,
>
> Thanks for the link. Based on the archive message the stable 3.5.X version
> will not be published within a month, and it is good to know cause we can
> design and plan better.
>
> Thanks,
> Joe
>
> On Fri, Mar 27, 2015 at 10:18 PM, Patrick Hunt <ph...@apache.org> wrote:
>
>> Hi Joe, have you seen this thread?
>> http://markmail.org/message/ymxliy2rrwjc2pmo
>>
>> Regards,
>>
>> Patrick
>>
>> On Thu, Mar 26, 2015 at 4:06 AM, Joe Meszaros
>> <jo...@impresstv.com> wrote:
>> > Hi,
>> >
>> > I've seen that you plan to make 3.5.1 RC soon, but it is not clear for me
>> > when do you plan to publish the 3.5.X stable version. I did not found
>> > usable information in the archived messages, nor on the internet.
>> >
>> > I am curious because we shipped a production system with Zookeeper and we
>> > will need the dynamic cluster reconfiguration.
>> >
>> > Many thanks,
>> > Joe
>>
>
>
>
> --
> *József **Mészáros*
> Senior Software Engineer
> +36 20 392 1243 <+36+20+392+1243>
> *impresstv.* <http://www.impresstv.com/>*com* <http://www.impresstv.com/>
Re: Zookeeper 3.5 stable
Posted by József Mészáros <jo...@impresstv.com>.
Hi Patrick,
Thanks for the link. Based on the archive message the stable 3.5.X version
will not be published within a month, and it is good to know cause we can
design and plan better.
Thanks,
Joe
On Fri, Mar 27, 2015 at 10:18 PM, Patrick Hunt <ph...@apache.org> wrote:
> Hi Joe, have you seen this thread?
> http://markmail.org/message/ymxliy2rrwjc2pmo
>
> Regards,
>
> Patrick
>
> On Thu, Mar 26, 2015 at 4:06 AM, Joe Meszaros
> <jo...@impresstv.com> wrote:
> > Hi,
> >
> > I've seen that you plan to make 3.5.1 RC soon, but it is not clear for me
> > when do you plan to publish the 3.5.X stable version. I did not found
> > usable information in the archived messages, nor on the internet.
> >
> > I am curious because we shipped a production system with Zookeeper and we
> > will need the dynamic cluster reconfiguration.
> >
> > Many thanks,
> > Joe
>
--
*József **Mészáros*
Senior Software Engineer
+36 20 392 1243 <+36+20+392+1243>
*impresstv.* <http://www.impresstv.com/>*com* <http://www.impresstv.com/>
Re: Zookeeper 3.5 stable
Posted by Patrick Hunt <ph...@apache.org>.
Hi Joe, have you seen this thread? http://markmail.org/message/ymxliy2rrwjc2pmo
Regards,
Patrick
On Thu, Mar 26, 2015 at 4:06 AM, Joe Meszaros
<jo...@impresstv.com> wrote:
> Hi,
>
> I've seen that you plan to make 3.5.1 RC soon, but it is not clear for me
> when do you plan to publish the 3.5.X stable version. I did not found
> usable information in the archived messages, nor on the internet.
>
> I am curious because we shipped a production system with Zookeeper and we
> will need the dynamic cluster reconfiguration.
>
> Many thanks,
> Joe