[
https://issues.apache.org/jira/browse/IGNITE-10663?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16984809#comment-16984809
]
Pavel Tupitsyn edited comment on IGNITE-10663 at 11/29/19 8:35 AM:
-------------------------------------------------------------------
[~avinogradov]
> I was pretty sure that C#/.NET parts can be delivered partially and on
> demand.
Sure, we could add WithReadRepair later. The problem is, it is now on public
API. If we did not catch the problem, it would end up in Ignite 2.8.0,
accessible to the users, but throwing a cryptic exception. Which would be a
shame, do you agree?
> Can we have some checklist or howTo for C#/.NET part to avoid such issues in
> future?
I've updated [Ignite.NET
Development|https://cwiki.apache.org/confluence/display/IGNITE/Ignite.NET+Development]
page on wiki with some guidelines on public API handling and tests.
Asking for a review is the best way to make sure the changes are good. Feel
free to ping me personally for anything C#/.NET related, you can use @ to
mention a user in JIRA or send a dev list email.
was (Author: ptupitsyn):
[~avinogradov]
> I was pretty sure that C#/.NET parts can be delivered partially and on
> demand.
Sure, we could add WithReadRepair later. The problem is, it is now on public
API. If we did not catch the problem, it would end up in Ignite 2.8.0,
accessible to the users, but throwing a cryptic exception. Which would be a
shame, do you agree?
I've updated [Ignite.NET
Development|https://cwiki.apache.org/confluence/display/IGNITE/Ignite.NET+Development]
page on wiki with some guidelines on public API handling and tests.
> Implement cache mode allows reads with consistency check and fix
> ----------------------------------------------------------------
>
> Key: IGNITE-10663
> URL: https://issues.apache.org/jira/browse/IGNITE-10663
> Project: Ignite
> Issue Type: Task
> Reporter: Anton Vinogradov
> Assignee: Anton Vinogradov
> Priority: Major
> Labels: iep-31
> Fix For: 2.8
>
> Time Spent: 20h
> Remaining Estimate: 0h
>
> The main idea is to provide special "read from cache" mode which will read a
> value from primary and all backups and will check that values are the same.
> In case values differ they should be fixed according to the appropriate
> strategy.
> ToDo list:
> 1) {{cache.withReadRepair().get(key)}} should guarantee values will be
> checked across the topology and fixed if necessary.
> 2) LWW (Last Write Wins) strategy should be used for validation.
> 3) Since LWW and any other strategy do not guarantee that the correct value
> will be chosen.
> We have to record the event contains all values and the chosen one.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
