Meng Xu
The current DD build all required teams in one big atomic block. When there are a lot of teams to build, say over 6000 teams, DD will take tens of...
### What would you like to be added/changed? The taint related test uses sleep to verify that the feature can replace pods on a timely manner. Using sleep() in unit...
Right now, `j tail` only shows the summary of the result. This makes debugging hard when the test is non-deterministic and hard to reproduce, such as rocksdb tests. One idea:...
I noticed that Joshua test has failure due to "Disk i/o error". This happens when Joshua agent process cannot write to its disk. This causes false-positive test failures (my guess)....
Since FDB can compile and run on ARM (https://github.com/apple/foundationdb/pull/2961), we may not want to break it accidentally. Based on the discussion in last week's community meeting, we may want to...
The performant restore (aka fast restore) performance can be limited by the destination FDB cluster's write throughput. To better understand the impact of the FDB cluster's write throughput on the...
The performant restore to be released in FDB 6.3 assumes that no restore worker can fail. This assumption is reasonable because (1) in a strictly controlled cluster, process and hardware...
The first version of the performant restore will be released in FDB 6.3 as an experimental feature. It thus closes the issue https://github.com/apple/foundationdb/issues/1049. The performant restore can still be further...
If backup files use a standard format, such as Protocol Buffers or FlatBuffers, digesting the backup will not need the verbose and error-prone deserialization process. It will make life easier...
If a storage server (SS) does not have a valid locality due to misconfiguration, the replication policy can select replicas that do not satisfy the policy. For example, in `three_data_hall`...