Re: Ignite Cluster going OOM
Hi, Vote for this release will be finished tomorrow. I think it will be successful and the release will be ready on this week. On Wed, Jul 26, 2017 at 12:57 PM, Ankit Singhai wrote: > Hi Andrew, > Please let us know when Ignite 2.1.0 would be available & do you think we > can try with Ignite 2.0? > > Thanks > Ankit Singhai > > > > > -- > View this message in context: http://apache-ignite-users. > 70518.x6.nabble.com/Ignite-Cluster-going-OOM-tp15175p15683.html > Sent from the Apache Ignite Users mailing list archive at Nabble.com. >
Re: Ignite Cluster going OOM
Hi Andrew, Please let us know when Ignite 2.1.0 would be available & do you think we can try with Ignite 2.0? Thanks Ankit Singhai -- View this message in context: http://apache-ignite-users.70518.x6.nabble.com/Ignite-Cluster-going-OOM-tp15175p15683.html Sent from the Apache Ignite Users mailing list archive at Nabble.com.
Re: Ignite Cluster going OOM
Thanks Andrew. Will wait few days for new release and after that would update you on the test results. -- View this message in context: http://apache-ignite-users.70518.x6.nabble.com/Ignite-Cluster-going-OOM-tp15175p15223.html Sent from the Apache Ignite Users mailing list archive at Nabble.com.
Re: Ignite Cluster going OOM
Looks weird. I see number of NODE_FAILED messages that means, some nodes doesn't respond for some reason. Possibly, Ignite can't release resources for such nodes during failure detection timeout. Try to decrease network timeout to allow failed nodes being dropped faster. Also you can try to switch to 2.1 version that should be available in next 3 days and will contains a lot of fixes. On Thu, Jul 20, 2017 at 7:10 PM, Ankit Singhai wrote: > ignite-d662243f.gz > <http://apache-ignite-users.70518.x6.nabble.com/file/ > n15188/ignite-d662243f.gz> > > It is server log file. > > Note:- 20 Caches are created but only 4 is being used. > > > > -- > View this message in context: http://apache-ignite-users. > 70518.x6.nabble.com/Ignite-Cluster-going-OOM-tp15175p15188.html > Sent from the Apache Ignite Users mailing list archive at Nabble.com. > -- Best regards, Andrey V. Mashenkov
Re: Ignite Cluster going OOM
ignite-d662243f.gz <http://apache-ignite-users.70518.x6.nabble.com/file/n15188/ignite-d662243f.gz> It is server log file. Note:- 20 Caches are created but only 4 is being used. -- View this message in context: http://apache-ignite-users.70518.x6.nabble.com/Ignite-Cluster-going-OOM-tp15175p15188.html Sent from the Apache Ignite Users mailing list archive at Nabble.com.
Re: Ignite Cluster going OOM
Load is 16K transactions / min it includes creation, updation, accessing & deleting it also. -- View this message in context: http://apache-ignite-users.70518.x6.nabble.com/Ignite-Cluster-going-OOM-tp15175p15187.html Sent from the Apache Ignite Users mailing list archive at Nabble.com.
Re: Ignite Cluster going OOM
Hi Andrew, Below is the grid config 10.179.29.135:47500..47509 10.179.29.136:47500..47509 10.179.29.137:47500..47509 I have not tried with less number of clients & increasing the number of servers. Please suggestion in this regard. Use case:- All the caches are storing session related data (such as login session and sso token management) which is created once, updating is not that frequent but accessing is 50 / 60 times. -- View this message in context: http://apache-ignite-users.70518.x6.nabble.com/Ignite-Cluster-going-OOM-tp15175p15186.html Sent from the Apache Ignite Users mailing list archive at Nabble.com.
Re: Ignite Cluster going OOM
Hi, Looks like a memory leak. Would you please share a grid config and clarify the use case? Does it happens with lower number of clients as well? Is increasing number of server nodes or heap size helpful? On Thu, Jul 20, 2017 at 3:20 PM, Ankit Singhai wrote: > Hi Guys, > I am using Ignite 1.8.0 with 3 servers and 30 clients. Where in each server > has 4 GB of java heap. Server is going OOM very soon and heap dump shows > the > below as top consumer. > > Class Name > | Shallow Heap | Retained Heap | Percentage > > > -- > org.apache.ignite.internal.processors.cache.distributed.dht. > GridDhtPartitionTopologyImpl > @ 0x705dd8970| 72 | 767,921,176 | 17.96% > org.apache.ignite.internal.processors.cache.distributed.dht. > GridDhtPartitionTopologyImpl > @ 0x700f38708| 72 | 749,744,568 | 17.53% > org.apache.ignite.internal.processors.cache.distributed.dht. > GridDhtPartitionTopologyImpl > @ 0x702763d90| 72 | 723,323,712 | 16.91% > org.apache.ignite.internal.processors.cache.distributed.dht. > GridDhtPartitionTopologyImpl > @ 0x702575918| 72 | 375,394,760 | 8.78% > > > -- > > > > I have 4 caches and they are configured to use OFFHEAP_TIERED > > Cache 1 > > > > > Cache2 > > > > > Cache3 > > > > > > Cache4 > > > > > Please help me out. > > > > -- > View this message in context: http://apache-ignite-users. > 70518.x6.nabble.com/Ignite-Cluster-going-OOM-tp15175.html > Sent from the Apache Ignite Users mailing list archive at Nabble.com. > -- Best regards, Andrey V. Mashenkov
Ignite Cluster going OOM
Hi Guys, I am using Ignite 1.8.0 with 3 servers and 30 clients. Where in each server has 4 GB of java heap. Server is going OOM very soon and heap dump shows the below as top consumer. Class Name | Shallow Heap | Retained Heap | Percentage -- org.apache.ignite.internal.processors.cache.distributed.dht.GridDhtPartitionTopologyImpl @ 0x705dd8970| 72 | 767,921,176 | 17.96% org.apache.ignite.internal.processors.cache.distributed.dht.GridDhtPartitionTopologyImpl @ 0x700f38708| 72 | 749,744,568 | 17.53% org.apache.ignite.internal.processors.cache.distributed.dht.GridDhtPartitionTopologyImpl @ 0x702763d90| 72 | 723,323,712 | 16.91% org.apache.ignite.internal.processors.cache.distributed.dht.GridDhtPartitionTopologyImpl @ 0x702575918| 72 | 375,394,760 | 8.78% -- I have 4 caches and they are configured to use OFFHEAP_TIERED Cache 1 Cache2 Cache3 Cache4 Please help me out. -- View this message in context: http://apache-ignite-users.70518.x6.nabble.com/Ignite-Cluster-going-OOM-tp15175.html Sent from the Apache Ignite Users mailing list archive at Nabble.com.