-
When rolling to a new metadata zone there is a significant amount write latency introduced.
The write latency is due to holding the file mutex(holding off any other metadata syncs and new file crea…
-
Hi, I was trying this command:
`sudo DEBUG_LEVEL=0 ROCKSDB_PLUGINS=zenfs make -j48 db_bench install`
But some errors occurred as below:
`$DEBUG_LEVEL is 0
Must specify package names on the command…
-
Background:
1. WAL sync should be enabled by default.
2. Multiple WAL files may be opened at the same time (only one WAI file could be written into)
3. Over 14 files could be concurrently opened &…
-
> Note: Please use Issues only for bug reports. For questions, discussions, feature requests, etc. post to dev group: https://groups.google.com/forum/#!forum/rocksdb or https://www.facebook.com/groups…
-
Add crash testing to zenfs, based on the rocksdb internal testing.
We can probably adapt/reuse what is available here:
- https://github.com/facebook/rocksdb/blob/master/tools/db_crashtest.py
-
`rocksdb.use_direct_io_for_flush_and_compaction` MTR test case run under ASan/Valgrind helped to reveal the following `heap-use-after-free` problem.
```
==527273== Invalid read of size 8
==527273==…
-
## TEST
DISK: ZNS * 2T
Rocksdb: 6.25
Zenfs: master
max_active_zones: 14
max_open_zones: 14
chunk_sectors: 4194304
chunk_sectors-size: 2M
## Problem
Two much amplification after fillrandom a…
-
BLKCLOSEZONE requires uapi headers from linux v5.5+.
The latest Ubuntu 20.04.3 LTS has only linux-libc-dev version (5.4.0-89.100).
IMHO it doesn't make sense to compile libzbd without BLKCLOSEZONE b…
-
For some reason, updates to header files require a clean rebuild to propagate. This can manifest in very hard-to-debug issues if one forgets to do a clean rebuild.
It looks like rocksdb is supporti…
-
```
zns@zns-server:~/rocksdb/plugin/zenfs/util$ make
g++ -std=c++11 -faligned-new -DHAVE_ALIGNED_NEW -DROCKSDB_PLATFORM_POSIX -DROCKSDB_LIB_IO_POSIX -DOS_LINUX -fno-builtin-memcmp -DROCKSDB_FALLOCAT…
Haltz updated
3 years ago