Re: H2database dependency in Apache Ignite

Hi, Ignite uses H2 as one of 2 available execution engines. Module 'ignite-indexing' depends on H2. Long story short, we can't bump H2 version anymore as some integration points were dropped in higher H2 versions. However, this CVE along with other known issues[1] do not affect the Ignite. [1]

Re: Fw:Expiry Policy can not working in gridgain 8.8.9

Hi, You wrote to the Apache Ignite user list. Apache Ignite and Gridgain are different projects with their own lifecycles and roadmaps. BTW, this looks like an issue in 8.8.9 that is already fixed in 8.8.18. Please contact the vendor for details.

Re: Ignite IPv6 support

Hi, As far as I know Ignite can't work with IPv6 on Mac devices. Also, we have a ticket related to K8s with IPv6. The ticket had a patch and likely being fixed in one of future releases. Please, find known issues in jira [1]. Ignite is not tested well in IPv6 environment and any help here will

Re: Failed to map SQL query

Hi, Most likely, the query intermediate result doesn't fit to JVM heap memory. The query may require all table data fetched before applying sorting. You can try to create a composite index over "act_id,mer_id,score" columns. On Tue, Aug 25, 2020 at 8:42 AM 1115098...@qq.com wrote: > Hi,an

Re: RE: Re: Unsafe usage of org.h2.util.JdbcUtils in Ignite

Hi, In Apache Ignite master branch I see a separate class H2JavaObjectSerializer that implements JavaObjectSerializer. Seems, this won't be released in 2.8 https://issues.apache.org/jira/browse/IGNITE-12609 On Thu, Mar 19, 2020 at 4:03 PM Andrey Davydov wrote: > Seem that refactor h2serilizer

Re: Geometry via SQL?

Hi, Please, refer to a Geo spatial index documentation [1] It is quite old, but should works. [1] https://apacheignite.readme.io/v1.7/docs/geospatial-queries#section-executing-geospatial-queries On Tue, Dec 10, 2019 at 10:57 PM richard.ows...@viasat.com < richard.ows...@viasat.com> wrote: > Is

Re: H2 version security concern

Hi, Mentioned CVE has no affect Ignite. Please, see discussion on dev-list. http://apache-ignite-developers.2346864.n4.nabble.com/H2-license-and-vulnerabilities-td40417.html#a40418 On Wed, Dec 11, 2019 at 2:22 AM Evgenii Zhuravlev wrote: > Hi, > > There are plans to replace H2 with Calcite.

Re: java.lang.OutOfMemoryError: GC overhead limit exceeded

Hi, Most likely heap size is too low. Try to increase Xmx up to 4Gb or higher or avoid G1GC usage on small heaps as it is very sensitive to free heap memory. Looks like you have Visor node (or may be web-console) in grid. Is OOM happened only when Visor attached to grid? On Fri, Mar 1, 2019 at

Re: about https://github.com/amsokol/ignite-go-client

Hi, Ignite has no module ignite-go-client. Client you mentioned is not maintained by Apache Ignite community. AFAIK, we have no plans to add GO client support for now. Feel free to contribute. Usually, Apache Ignite is released twice a year. The latest version has been released at December

Re: Text Query question

Hi, Unfortunatelly, it doesn't look as an open source solution. It is not clear how their Indices are integrated with Ignite page memory. If they do not use Ignite page memory then how they survive in failover scenarios as no shared test\test results are available? Otherwise, I bet they have

Re: Migrate from 2.6 to 2.7

Hi Andrey, It looks like you try to run "Select" sql query inside explicit transaction. Please, let us know if it is not true. This was workable in 2.7 as SQL had no transactional support and query just ignored transactional context (however "Select for Update" wasn't). For now, SQL Select query

Re: SQL Query plan confusion

Hi, The difference in query plans is a 'bad' one use PUBLIC.TEST_DATA_K_V_ID index instead of PUBLIC.TEST_DATA_ID_K_V. Have you tries to remove other indices and left TEST_DATA_ID_K_V only? It looks weird for me, the 'good' query plan shows unicast requests as ID is affinity field, but 'bad'

Re: Slow select distinct query on primary key

Yuri, how did you get inline size 60? I'd think 55 should be enough to inline Account_ID. 55 = 1 /* byte, type code */ + 4 /* int, array lenght */ + 50 /* data size for ANSI chars */ On Fri, Nov 30, 2018 at 1:25 PM Юрий wrote: > Please provide explain plan of the query to check that index is

Re: Cache Structure

Hi, Yes, to add a new item to a list value of second cache Ignite will have to deserialize whole list (with all it's items) then add new item and then serialize list again. You can try to use BinaryObjects to avoid unnecessary deserialization of list items [1]. Also note, k2 and k1 data will have

Re: How could imporeve UNION ALL performance

Hi, Can you share a query plan? Do you have any indices on last and lastprice columns? Have you tried to run queries separately without Union? чт, 8 нояб. 2018 г., 23:47 wengyao04 wengya...@gmail.com: > Hi we use ignite as our searching backend. > Our cache is in replicated mode and

Re: OVER() clause in ignite

Hi, Ignite have no support window functions due to lack of underlying H2. OVER is a part of SQL ANSI'03 standart, but Ignite claims to be ANSI'99 complaint. The only workaround is to rewrite query with using joins or calculate result manually. User aggregate functions also are not supported and

Re: Ignite Query Slow

Hi, Ignite provides 2 way for configuring indices: queryEntity and annotation based. Seems, you either forget to setIndesTypes or tring mix both of these approaches. Please, find examples in documentation available via link I've provided earlier. пн, 1 окт. 2018 г., 18:05 Skollur : > I tried

Re: Ignite Query Slow

Please take a look at this. https://apacheignite.readme.io/v2.6/docs/indexes#section-queryentity-based-configuration 29 сент. 2018 г. 3:41 пользователь "Skollur" написал: Thank you for suggestion. Can you give example how to create secondary indices? -- Sent from:

Re: Apache Ingite Join returns wrong records

Hi, Try to use qry.setDistributedJoins(true). This should always return correct result. However, it has poor performance as due to intensive data exchange between nodes. By default, Ignite join only the data available locally on each node. Try to collocate your data to get better performance

Re: Ignite Query Slow

Hi, Please, try to create secondary indices on join columns, otherwise query will fall into full table scan. Then if you will see SCANs, as next step, you can try to rewrite your query with different tables join order. Sometimes, underline H2 can change join order to non optimal. In that case

Re: Query execution too long even after providing index

HI, Have you tried to increase index inlineSize? It is 10 bytes by default. Your indices uses simple value types (Java primitives) and all columns can be easily inlined. It should be enough to increase inlineSize up to 32 bytes (3 longs + 1 int = 3*(8 /*long*/ + 1/*type code*/) + (4/*int*/ +

Re: Query 3x slower with index

Hi Actually, first query uses index on affinity key which looks more efficient than index on category_id column. The first query can process groups one by one and stream partial results from map phase to reduce phase as it use sorted index lookup, while second query should process full dataset on

Re: Slow SQL query uses only a single CPU

m SQL? >> >> From your description, it seems like the query is executed in the >> following order >> 1) Group by customer_id >> 2) For each group, perform the filtering on date using the index and >> aggregates >> >> My impressions was that the

Re: Slow SQL query uses only a single CPU

y to check what Ignite is actually doing? How are > indexs used (by Ignite or H2)? > > Cheers, > Eugene > > On Wed, Aug 22, 2018 at 3:54 AM, Andrey Mashenkov < > andrey.mashen...@gmail.com> wrote: > >> Hi, >> >> 1. Possible there are too much dat

Re: In the case of read-through cache, both Replicated and Partitioned mode are identical?

Hi, CacheStore semantic suppose you will warm-up a cache via calling cache.loadCache(). When you use read-through and forget load cache from back store then it is ok if no backups are updated on read operation. Backups and replicated cache in case of read-through just add some tolerance to

Re: Slow SQL query uses only a single CPU

Hi, 1. Possible there are too much data should be looked for the query. With single node and parallelism=1 query will always run in single thread. You can try to add more nodes or increase query parallelism to utilize more CPU cores. 2. Index on date field may be not effective as reduce phase

Re: Query execution too slow

Hi, Composite index field order matters. Try to set order=3 for "value" column. On Tue, Aug 21, 2018 at 10:43 PM Ilya Kasnacheev wrote: > Hello! > > Yes, I am afraid that you will need another index. > > Regards, > > -- > Ilya Kasnacheev > > 2018-08-21 19:55 GMT+03:00 Prasad Bhalerao : > >>

Re: And again... Failed to get page IO instance (page content is corrupted)

ked the code to get rid > of expiration policy. > All our non-persistent caches have expiration policy but this should not > be a problem, right? > > BR, Oleksandr > > On Thu, Jun 28, 2018 at 8:37 PM, Andrey Mashenkov < > andrey.mashen...@gmail.com> wrote: > >> Hi

Re: And again... Failed to get page IO instance (page content is corrupted)

uses a catastrophe. > This mean no availability in fact - I had to clean data folders to start > my cluster after that > > BR, Oleksandr > > > On Fri, Jun 22, 2018 at 4:06 PM, Andrey Mashenkov < > andrey.mashen...@gmail.com> wrote: > >> Hi, >> >> We

Re: Ignite 2.4 & 2.5 together?

Hi, It is recommended to set baseline at first if grid was upgraded from <2.4 version. Then you need to deactivate the grid to force checkpoint, this can reduce time of next startup. Then stop grid and update jars to newer version and start the grid. Then you may be need to activate grid if it

Re: Ignite 2.4 & 2.5 together?

Hi, It should not be possible by default and highly not recommended. As there can be differences in communication protocol between versions ans compatibility is not tested. However, persistence format should be backward compatible that means Ignite newer version can work with persistence created

Re: And again... Failed to get page IO instance (page content is corrupted)

Hi, We've found and fixed few issues related to ExpiryPolicy usage. Most likely, your issue is [1] and it is planned to ignite 2.6 release. [1] https://issues.apache.org/jira/browse/IGNITE-8659 On Fri, Jun 22, 2018 at 8:43 AM Olexandr K wrote: > Hi Team, > > Issue is still there in 2.5.0 > >

Re: Quick questions on Evictions

Yes, you are right. On Wed, Jun 20, 2018 at 4:13 PM the_palakkaran wrote: > So to conclude, if I have enabled on heap storage for cache(using > cache.setOnHeapEnabled(true), > then : > 1. Still data will be stored off heap, but will be loaded to heap. To > escape > out of memory error, I have

Re: Quick questions on Evictions

DataRegionConfiguration.setMaxSize() should be used to limit offheap memory usage. On Tue, Jun 19, 2018 at 2:35 PM the_palakkaran wrote: > So how do I limit cache size if ignite native persistence is enabled using > dataRegionCfg.setPersistenceEnabled(true)? I don't want it to keep a lot of >

Re: SQL Query full table scan, Node goes down

Hi, Try to set Lazy flag for SQL query [1]. [1] https://apacheignite-sql.readme.io/docs/performance-and-debugging#result-set-lazy-load On Tue, Jun 19, 2018 at 8:06 PM bhaskar wrote: > Hi, > I have Ignite 5 node cluster with more than dozen tables(Cache) . Our > client > are using SQL and

Re: Quick questions on Evictions

Hi, DataPageEvictionMode is about algorithm of choosing page to be replaced. EvictionPolicy is what you are looking for. E.g. FifoEvictionPolicy or LruEvictionPolicy. It looks like EvictionPolicies can't be used with persistence as all of them uses non-persistent structures to track cache

Re: Error while Starting Grid: javax.management.InstanceAlreadyExistsException: (LruEvictionPolicy)

t; > cc.setEvictionPolicy(evpl); > > cc.setExpiryPolicyFactory(CreatedExpiryPolicy.*factoryOf*( > *new *Duration(TimeUnit.*SECONDS*, 15))); > > cc.setStatisticsEnabled(*true*); > > > > > > > > *From:* Andrey Mashenkov [mailto:andrey.mashen...@gmail.com] > *Sent:*

Re: MSSQL as backend for cache

Hi, SAXParseException says a namespace was missed. Hope this [1] will helpful. [1] https://stackoverflow.com/questions/2897819/spring-using-static-final-fields-constants-for-bean-initialization On Wed, Jun 13, 2018 at 1:19 PM, Michaelikus wrote: > it seems error was because of declaration of

Re: And again... Failed to get page IO instance (page content is corrupted)

Hi, Possibly, it is a bug in partition eviction optimization. Ignite can skip partition eviction procedure and remove partition instantly if there is no indexes. If it is so, you can try the latest ignite-2.5 version [1] or as a workaround you can add index via configuring QueryEntity or via

Re: MSSQL as backend for cache

Hi, You've changed invalid URL property to another invalid one. Anyway, I see no Ignite issues here. Please, take a look how to configure MsSQL JDBC driver [1]. [1] https://stackoverflow.com/questions/11206673/ms-sql-configuration-in-beans-xml On Wed, Jun 13, 2018 at 10:50 AM, Michaelikus

Re: RDBMS as backend for Ignite SQL

Hi, You can use only one backed store per cache. I'm not sure, but very likely you will not be able to use cross-cache transaction for caches backed by different stores. Anyway, you can try. On Sat, Jun 9, 2018 at 1:15 PM, Michaelikus wrote: > Is it possible to use various RDBMS as persisted

Re: MSSQL as backend for cache

Hi, Looks like it is MsSql driver issue: Error setting property values; nested exception is org.springframework.beans.NotWritablePropertyException: Invalid property 'URL' of bean class [com.microsoft.sqlserver.jdbc.SQLServerDriver]: Seemd, URL is generated by driver itself automatically. Try

Re: Failsafe re-entrant locks still held after a node exits.

No. I see no message. On Fri, Jun 8, 2018 at 7:27 PM, Jon Tricker wrote: > Do you see the message "ERROR ; Lock is not released on exit" printed by > the code? > > > I believe if things are working correctly it should not be. > > > -------

Re: How are SQL Queries executed on Ignite Cluster

1. It is not supported for now, but we plan do fix it [1] 2. Not yet. We split index different way. Every tree can manage certain partition numbers. Feel the difference. Mapping "partition -> tree segment" is static and we use as reminder of the division, like partition_id%N. Query works same way

Re: Ignite 2.5 | Can't restore memory - critical part of WAL archive is missing with walMode=NONE

Hi Emmanuel, Sorry for late answer. I've found I check your test against master branch and test passes, but it fails on 2.5 branch. It looks like already fixed. On Wed, Jun 6, 2018 at 3:47 PM, Emmanuel Marchand < emmanuel.march...@exensa.com> wrote: > I was wrong on the introduction of the

Re: Node pause for no obvious reason

Possibly there is a race. I've created a ticket for this [1] [1] https://issues.apache.org/jira/browse/IGNITE-8751 On Fri, Jun 8, 2018 at 4:56 PM, Andrey Mashenkov wrote: > Hi, > > Looks like node was segmented due to long JVM pause. > There are 2 "long JVM pause&

Re: Error while Starting Grid: javax.management.InstanceAlreadyExistsException: (LruEvictionPolicy)

Hi, Looks like a bug. Can you share grid configuration? Do you have more than one node in same JVM? Do you have configure cache groups manually? On Fri, Jun 8, 2018 at 4:48 PM, HEWA WIDANA GAMAGE, SUBASH < subash.hewawidanagam...@fmr.com> wrote: > Hi everyone, > > > > As a quick note on what

Re: Node pause for no obvious reason

Hi, Looks like node was segmented due to long JVM pause. There are 2 "long JVM pause" messages in long an suspicious long checkpoint: Checkpoint finished [cpId=77cf2fa2-2a9f-48ea-bdeb-dda81b15dac1, pages=2050858, markPos=FileWALPointer [idx=2051, fileOff=38583904, len=15981],

Re: Ignite C# Geometry

Hi, Here is an example hot to create binary object [1]. I'm not sure it will work, but you can try to pass full class name as BinaroObject type. Smth like that: IBinaryObjectBuilder builder = ignite.GetBinary().GetBuilder("com.vividsolutions..."); [1]

Re: Failsafe re-entrant locks still held after a node exits.

Hi, I've run your code several times and see no errors. On Fri, Jun 8, 2018 at 12:59 PM, Jon Tricker wrote: > Hi. I have been observing a problem on 2.5.0 where a node shuts down but > locks held by it appear to still be held. > > > > Hopefully the attached demo code illustrates the issues

Re: Ignite benchmarking on K8s cluster

Hi, What's actually wrong with yardstick benchmarks? On Thu, May 24, 2018 at 10:34 AM, vbm wrote: > Hi, > > We are trying to use yardstick benchmark from gridgain > (https://www.gridgain.com/resources/benchmarks/running-gridgain-benchmarks > ). > These tests are mainly for a non K8s cluster.

Re: IgniteCache invoke CacheEntryProcessor, not throw EntryProcessorException

Hi, Yes, but no. You will observe lower latency per operation on client and you will be able to utilize more server resources with less number of threads. But operation logic will be the same, no steps will be omit (backups will be updated as well)... so, maximal performance that can be achieved

Re: ignite peer class loading query

1. Peer-classloading is designed for compute jobs. All other classes (except may be user domain classes when BinaryObjects are used) must be in classpath on all nodes. 2. Next is true for transactional cache. Client node must have a CacheStore drivers as well as server nodes. This because a

Re: Cache getting cleareing automatically

Ignite should support persistence format of older versions. Also, please, take a look at Baseline topology feature that is available from 2.4 version. [1] https://apacheignite.readme.io/docs/baseline-topology On Thu, Jun 7, 2018 at 5:35 PM, siva wrote: > If we switch to 2.5 version ,what

Re: Cache getting cleareing automatically

Hi, Try to switch to the latest Ignite 2.5 version as it includes many bugfixes related to cluster stability. On Thu, Jun 7, 2018 at 4:57 PM, siva wrote: > 1.we are using ignite 2.3 version , > 2.We are using an Api.Client node is a spring boot rest application ,we > have > written an api to

Re: Does equals method have relevance in Custom Key Class

Yes. there are scenarios you can't use simple data types. This relates to not only Ignite. As Ignite SQL layer is built over key-value storage, there always be some overhead on SQL operations. Key-value API allow you operate with one or more cache entries with known keys. If the keys of entries

Re: Cache getting cleareing automatically

Hi, What Ignite version do you use? Looks like it was fixed in 2.5 [1]. How are you create and destroy a cache? via API or SQL? [1] https://issues.apache.org/jira/browse/IGNITE-8021 On Thu, Jun 7, 2018 at 4:32 PM, siva wrote: > > Hi, > We have two ignite server nodes in single physicle

Re: Ignite 2.5 nodes do not rejoin the cluster after restart (works on 2.4)

Hi, What baseline topology does ./control.sh prints? Is it possible, a node that out of baseline has started before baseline node starts? On Thu, Jun 7, 2018 at 9:54 AM, szj wrote: > Well, it definitely does work in 2.4. Please notice that there needs to be > ignitevisorcmd.sh involved to

Re: Ignite C# Geometry

Hi, I don't think JST classe are supported in Ignite binary protocol. You can try to use BinaryObjects for mock vividsolution classes or inline in SQL query string. On Thu, Jun 7, 2018 at 10:34 AM, Jonathan Mayer wrote: > Hi Guys, > > I realise that 2.5 has just been released and that

Re: Ignite Node failure - Node out of topology (SEGMENTED)

Hi, Seems, there is a bug with IPv6 usage [1]. This has to be investigated. Also, there is a discussion [2]. [1] https://issues.apache.org/jira/browse/IGNITE-6503 [2] http://apache-ignite-developers.2346864.n4.nabble.com/Issues-if-Djava-net-preferIPv4Stack-true-is-not-set-td22372.html On Wed,

Re: IgniteCache invoke CacheEntryProcessor, not throw EntryProcessorException

Hi, Looks like it is ok. Ignite have next sync modes full_sync, primary_sync, full_Async. You can detect failure from user thread if it happens on synchronous phase. FULL_SYNC: cache operation return control to user thread after operation has been applied on primary and all backups. All

Re: Ignite Cluster getting stuck when new node Join or release

Hi, It is ok if you kill client node. Grid will wait for failureDetectionTimeout before drop failed node from topology. All topology operations will stuck during that time as ignite nodes will wait for answer from failed node until they detected failure. On Thu, Jun 7, 2018 at 8:22 AM, Sambhaji

Re: ClassCastException When Using CacheEntryProcessor in StreamVisitor

Hi, Is it possible you change lambdas code between calls? Or may be classes are differs on nodes? Try to replace lambdas with static classes in your code. Will it work for you? On Tue, Jun 5, 2018 at 10:28 PM, Cong Guo wrote: > Hi, > > > > The stacktrace is as follows. Do I use the

Re: IgniteCache invoke CacheEntryProcessor, not throw EntryProcessorException

Hi, Would you please share a full config? or reproducer if possible? On Tue, Jun 5, 2018 at 9:56 PM, haotian.chen wrote: > Hi Developers, > > I am not sure if I understand the IgniteCache invoke method API correctly > here: > https://ignite.apache.org/releases/latest/javadoc/org/ >

Re: Does equals method have relevance in Custom Key Class

Hi, 1. Yes, use simple types if it is possible. These types has much smaller footprint than POJOs. Ignite has performance optimization for simple types (and some other JDK classes like String). Ignite use BinaryObjects [1] concept that allows to query certain user-object's field without

Re: Ignite 2.5 nodes do not rejoin the cluster after restart

HI, Is it possible there are nodes out of baseline started ans node with baseline is able to discover them? On Wed, Jun 6, 2018 at 9:48 AM, szj wrote: > I also tested an upgrade of the PoC 2-node cluster running Ignite 2.4 to > 2.5. > Both nodes shut down, upgraded, started on node1, started

Re: How Ignite performs cache.get() operation when hashCode is overridden ?

Duplicate. Answeder on SO [1] [1] https://stackoverflow.com/questions/50714462/how-ignite-performs-cache-get-operation-when-hashcode-is-overridden On Wed, Jun 6, 2018 at 10:03 AM, the_palakkaran wrote: > > > > See the

Re: Does equals method have relevance in Custom Key Class

Hi, 1. There are many ways to improve SQL query performance. The basic rules are to use simple types (primitives, String, UUID and some other JDK classes), use inices with proper inline size to avoid unnecessary row lookups, use right collocation. 2. This will not work. Ignite fully relies on

Re: Ignite client mvn artifact's purpose?

Hi, Answered on SO, please take a look [1]. [1] https://stackoverflow.com/questions/50691453/ignite-client-mvn-artifacts-purpose On Sun, Jun 3, 2018 at 2:26 PM, mlekshma wrote: > Hi Folks, > > I notice there is a mvn artifact with artifact id "ignite-clients". What is > its purpose? > >

Re: support for Scala Collection as key?

Hi, Most likely, Scala Vector can't be used as a key as-is due to it's implementation [1]. I'm not sure, if k1 and k2 scala objects have same internal structure to Ignite can serialize them to same binary representation. [1]

Re: Doubts regarding using externalizable in model classes

Hi, It is a duplicate of StackOverflow question [1]. Please, find an answer on SO via link provided. [1] https://stackoverflow.com/questions/50697634/using-externalizable-in-apache-ignite On Tue, Jun 5, 2018 at 8:30 AM, the_palakkaran wrote: > I have 3 doubts, > > 1. My model classes are

Re: Ignite 2.5 | Can't restore memory - critical part of WAL archive is missing with walMode=NONE

pty), the activation completes successfully. > Regards, > --- > Emmanuel. > > > > On 05/06/18 15:40, Andrey Mashenkov wrote: > > Hi, > > I can't reproduce the issue. > Is it possible grid configuration was changed between runs? > Is it possible to share a reproducer? >

Re: support for Scala Collection as key?

Hi, Ignite doesn't rely on BinaryObject hashcode or equals method, it use internal comparison logic instead. So, it is ok BinaryObject.equals() return true if compared with same object only. As Ignite have no hooks for scala collections, they will be handled as regular user objects. Let us know

Re: Updating the same entry multiple time in Ignite transaction

he, I get transaction > optimistic exception. > But instead of submitting the task, if I just call task.run() method it > works fine. > > > Thanks, > Prasad > > On Tue, Jun 5, 2018 at 7:07 PM, Andrey Mashenkov < > andrey.mashen...@gmail.com> wrote: > >> Hi, >

Re: Ignite 2.5 | Can't restore memory - critical part of WAL archive is missing with walMode=NONE

Hi, I can't reproduce the issue. Is it possible grid configuration was changed between runs? Is it possible to share a reproducer? On Tue, Jun 5, 2018 at 12:05 PM, Emmanuel Marchand < emmanuel.march...@exensa.com> wrote: > Hi, > > I'm testing v2.5 vs v2.4 for persisted dataregion with

Re: Updating the same entry multiple time in Ignite transaction

t I don't understand why transaction fails if the entry is modified > after read operation in same transaction? > > Thanks, > Prasad > > On Tue, Jun 5, 2018 at 6:07 PM, Andrey Mashenkov < > andrey.mashen...@gmail.com> wrote: > >> Hi, >> >> It is ok

Re: Ignite Optimistic Transactions

Hi, 1. The explanation looks confusing. In other words, Ignite remembers entry version in transaction context at first read access. 2. If another transaction has commited just before the current transaction, then current transaction will fails with TransactionOptimisticException as Ignite can't

Re: How to execute an sql on a specific remote server?

Hi, What are you trying to achieve? What is cache mode? For REPLICATED cache, your query will be sent to and executed on one random node. 1. You can sent a compute job [1] to certain node (via ignite.compute(ignite.cluster().forNode(...)).execute(task)) or affinity job [2]

Re: Updating the same entry multiple time in Ignite transaction

xt.commitTxAsync(GridCacheSharedContext.java:975) > at org.apache.ignite.internal.processors.cache.transactions. > TransactionProxyImpl.commit(TransactionProxyImpl.java:288) > ... 10 common frames omitted > > Thanks, > Prasad > > On Tue, Jun 5, 2018 at 2:1

Re: Updating the same entry multiple time in Ignite transaction

Hi, Sure, multiple actions with same query within same transaction should work. Please, let us know if you observe unexpected behavior. Any reproducer will be appreciated. On Tue, Jun 5, 2018 at 10:36 AM, Prasad Bhalerao < prasadbhalerao1...@gmail.com> wrote: > Hi, > > Can I update the same

Re: How to use Binarylizable interface or Externalizable on my custom Key?

Hi, You are free to use any hashcode and equals in your classes. Ignite will convert your POJO to BinaryObject before save to cache and use it's own hashing anyway. It still unclear what are trying to do. Key is unique for each cache entry. So, it is impossible to get entry by a part of key. You

Re: How to use Binarylizable interface or Externalizable on my custom Key?

Hi, Would you please describe your scenario? Why you need to use binarylizable or Externalizable? Ignite allows overriding methods, but ignores object's hashcode and equals and rely on its own implementation. Ignite operates with BinaryObjects underneath and use BinaryObject hash codes. This

Re: IGNITE performance very very very less !!!

Hi, Why you think the performance is slow? Ignite SQL has an H2 underneath and have some overhead as it is a distributed system at first. So, local query latency can be slightly higher. How do you measure query latency? You may need to warm grid up to all components involved in query initialize.

Re: Cache not rebalanced after one node is restarted

Hi, It is possible, you put too few data that belongs to 2 partitions (total partition is 1024 by default) and after restart you got different partition distribution for 2 nodes. Please take a look at how affinity function works [1]. [1]

Re: read through cache refresh

Hi, When your data in back store has been changed, you should clear cache to force ignite forget outdated data. Yes cache.clear() will be enough. If you find this method slow and you have a large dataset, then you can try to destroy and recreate a cache instead. Cache will not blocked during

Re: Ignite Node failure - Node out of topology (SEGMENTED)

Hi, Try to disable IPv6 on all nodes via JVM option -Djava.net.preferIPv4Stack=true [1] as using both IPv4 and IPv6 can cause grid segmentation. [1] https://stackoverflow.com/questions/11850655/how-can-i-disable-ipv6-stack-use-for-ipv4-ips-on-jre On Fri, Apr 27, 2018 at 8:52 AM, naresh.goty

Re: Hibernate 5.2 support

Hi, As ticket still has no patch, most likely it will be postponed to the next version. On Mon, Apr 16, 2018 at 4:27 PM, kestas wrote: > I see there is ticket registered for Hibernate 5.2 support: > https://issues.apache.org/jira/browse/IGNITE-5848 > It is planned for 2.5

Re: NullPointerException in GridCacheTtlManager.expire

Hi Dome, It is known issue and there is a ticket for this [1] you can track. It is hard to be reproduced due to a race. Seems, GridCacheUtils.unwindEvicts() should check if context has been started. [1] https://issues.apache.org/jira/browse/IGNITE-7972 On Thu, Apr 12, 2018 at 2:39 PM,

Re: Running heavy queries on Ignite cluster on backing store directly without impacting the cluster

Hi, >So are you saying, query execution will not have any impact on the cluster >activities like GET/PUTs in general ?? Why? Queries and get\put share same resources. So, they will affect each other. >I was thinking, if we can run these queries directly on backing store, it >may not have any

Re: How to insert multiple rows/data into Cache once

s write through approach work with jdbc bulk update/insert/ delete? >> >> >> 4) Does ignite have any apis on cache only for update purpose? Put/putAll >> will insert or overwrite. What if I just want to update existing entries ? >> >> >> Thanks, >>

Re: Using 3rd party DB together with native persistence (WAS: GettingInvalid state exception when Persistance is enabled.)

Hi It is very easy for user to shoot himself in a foot... and they do it again and again e.g. like this one [1]. [1] http://apache-ignite-users.70518.x6.nabble.com/Data-Loss-while-upgrading-custom-jar-from-old-jar-in-server-and-client-nodes-td20505.html On Thu, Mar 8, 2018 at 11:28 PM,

Re: Any idea what this is ?

t>(BinaryReaderExImpl.java:162) ~[ignite-core-2.4.0.jar:2.4.0] > at > org.apache.ignite.internal.binary.GridBinaryMarshaller.deserialize(GridBinaryMarshaller.java:310) > ~[ignite-core-2.4.0.jar:2.4.0] > at org.apache.ignite.internal.binary.BinaryMarshaller. > unmarshal0(B

Re: Performance of Ignite integrating with PostgreSQL

Hi, 1. Have you tried to test your disk write speed? it is possible it was formatted with no align respect. 2. Have you tried to check if there is any Postgres issues? E.g. via querying postgres directly. Do you see higher disk pressure? 3. Is it possible you generate data too slowly? Have you

Re: Spark 'close' API call hangs within ignite service grid

Hi, Socket exception can be caused by wrong network configuration or firewall configuration. If node will not be able to send an response to other node then it can cause grid hanging. (Un)marshall exceptions (that is not caused by network exceptions) is a signal that smth goes wrong. On Fri,

Re: Running heavy queries on Ignite cluster on backing store directly without impacting the cluster

Hi, >In this case also, since we dont have eviction in place, all the time data >is retrieved from RAM only, the only time request goes to Oracle is for >Upserts and delete. Here, all queries run on data in RAM only. Ignite just propagate updates to back store (to Oracle via CacheStore impl) with

Re: Ignite Client Heap out of Memory issue

idn't. For storm worker, it behaves like a normal fatal error > which make storm worker restart. > 2) It did make ignite server heap dump. for I analyzed the dump, it is > ignite server's. (seeing from PID and java command etc.) > > anyone can explain? Thanks. > > Thanks > Shawn >

Re: Distributed lock

Hi, If lock will be resides on node that hold the lock, how newly joined node will discover just locked lock instance by lock name to add itself in waiting queue? Who will be responsible for waiting queue handing? Does the queue should be transferred when owner has changed? And finally, how this

Re: Slow data load in ignite from S3

the frequency, we increased the buffer size , but > it didn’t have any impact on performance > > On Fri, 30 Mar 2018 at 10:49 PM, Andrey Mashenkov < > andrey.mashen...@gmail.com> wrote: > >> Hi, >> >> Possibly, storage is a bottleneck or checkpoint buffer is too la

Re: How to insert multiple rows/data into Cache once

; > > By the way, I have tried the [1], that worked well. > > > > Finally, I still need to use the SQL as a client node, and quick write > data into cache. > > > > Thank you for helping me > > > > Rick > > > > > > *From:* Andrey Mashenkov [mai

Re: Ignite Client Heap out of Memory issue

Hi Shawn, 1. Ignite use off heap to store cache entries. Client store no cache data. Cache in LOCAL mode can be used on client side, and it should use offheap of course. All data client retreive from server will be in offheap 2. It is not IgniteOutOfMemory error, but JVM OOM. So, try to

  1   2   3   4   5   6   >