-
In [the Naiad paper](https://dl.acm.org/citation.cfm?id=2522738), we learn that:
> Naiad has a simple but extensible implementation of fault tolerance: each stateful vertex implements a `CHECKPOINT…
-
Hello!
As far as I can see when the OutboxDispatcherService encounters any exception (db or Kafka may be temporary unavailable) it stops the message processing cycle without any obvious way to recove…
-
## Background information
### What version of Open MPI are you using? (e.g., v4.1.6, v5.0.1, git branch name and hash, etc.)
v5.0.1
### Describe how Open MPI was installed (e.g., from a sourc…
kwnsl updated
7 months ago
-
In order to test correctness and robustness of our framework implementation under failure, we need to able to simulate the node failure and communication failure. For now we can start with simulate th…
-
While working on Quarkus Tools for IJ, I noticed the following properties:
![image](https://user-images.githubusercontent.com/695993/136550388-c73dd04f-1723-46f5-bf0f-89960f7f24b3.png)
The same …
-
Section 5.2 of the Chord paper describes a mechanism for making Chord more fault tolerant by keeping a list of log(N) successors at each node (and replicating keys you own at those nodes). We should h…
-
Currently fault tolerant VMs will essentially show up as two VMs with the same information. This causes a few duplicates whenever these VMs show up on report, as well as causing false positives for th…
-
In the paper, it says that "Thrill currently provides no fault tolerance. While our data-flow API permits smooth integration of fault tolerance using asynchronous checkpoints the execution model of ex…
-
### Fault tolerance to dead actors
The documentation on actors fault tolerance is not clear whether the actors a re-created: https://ray.readthedocs.io/en/latest/fault-tolerance.html
The line ment…
-
Jira Link: [DB-2359](https://yugabyte.atlassian.net/browse/DB-2359)
From https://phabricator.dev.yugabyte.com/D8253#243515:
"
We should also consider some unhappy paths:
- crash or leader failover…