onlyjob

FTBFS: dns/dns.h:1009:24: error: lvalue required as unary '&' operand

2

Encountered the following FTBFS on Debian "unstable" _x86_64_: ~~~~ In file included from ip.c:44: ip.c: In function 'mill_ipremote_': dns/dns.h:1009:24: error: lvalue required as unary '&' operand 1009 | #define dns_opts(...)...

safeguard against reuse of active caching device

9

I've managed to wreck dm-writeboost module by accidentally re-using active caching device by constructing another dm-writeboosted device using the same caching device. dm-writeboost went into endless loop constantly reading from...

please provide man page; use POD documentation

12

Please introduce man page and POD documentation. Man page can be conveniently generated from POD markup using [Pod::Usage](http://perldoc.perl.org/Pod/Usage.html) core module. Thanks.

[Currently used](https://github.com/gravitational/teleconsole/blob/master/vendor/github.com/Sirupsen/logrus/CHANGELOG.md) `github.com/sirupsen/logrus` 0.10.0 is from early 2016 -- too old. Please consider upgrading to more recent release: https://github.com/sirupsen/logrus/releases

[DEFAULT] Let's use `ATIME_MODE = 2` (relatime) by default

8

I have been using MooseFS (and LizardFS before it) for a while and I have years of operational experience of managing 150+ TiB cluster running variety of application and `/home`...

chunkserver: workers should have timeouts? [errors handlling]

5

I did something silly that exposed another problem in chunkservers: I've activated a new _ext4_ feature - `fast_commit` that promises significant improvements to _fsync_ times. (`tune2fs -O fast_commit {DISK}`). On...

[FEATURE] notify support

1

It would be awesome to implement cross-client _notify_ support (that currently works only within particular mount). Apparently there is [Inotify support in FUSE and virtiofs](https://lwn.net/Articles/874000/). See also [Change notifications for...

I snapshot millions of files/directories daily and the size of snapshot is such that removing it takes hours... LizardFS have a recursive directory removal tool (invoked as `lizardfs rremove`) that...

CGI: incorrect number of chunk replications

8

"Server Charts" --> "number of chunk replications per minute" incorrectly show number of _attempted_ replications instead of actually _completed_ replications. On a particular chunkserver free space was exhausted and CGI...

[BUG] chunkserver: catch ext4 read-only state

7

While number of chunks on chunkserver did not change, I've noticed unrealistically high number of ongoing chunks deletions (13k) in chart, which reminded me of #273 (don't make fake charts...