@commdserv

0 Followers
30 Following
29 Posts
@lina Well I ran into some tech issues today, and bing seldom got me to actual solution... like sddm autologin not working on opensuse(this was due to a sysconfig), or force edid failed to mock display(this one due to kernel cmdline order). Even llms do better job, especially searching semantically and get most relevant results compared to bing/ddg
@lina don't use google. you may try adding more keywords. also I find ai sometimes searches better keywords than myself, perhaps bcuz I tell it more details🤣
@yuuta 老大你的channel没了喵
@felixonmars 至少比simd.cc违规能解释)
@lispi314 In this case bees the dedup tool was doing an initial dedup with lots of snapshots and tight space left... Maybe just got used too hard😇
auto bg reclaim can help, but at that time it was off by default. Don't know if it's default on now.
@lina https://blog.lilydjwg.me/2023/7/6/a-btrfs-incident.216658.html
Bees dedup ate up all metadata space, and no space for new block group. Deleting snapshot also used up global reserve, so the commit stuck... and forced ro
btrfs 翻车记 - 依云's Blog

btrfs 翻车记- 如标题所言,我用了多年的 btrfs,终于还是遇到翻车啦。由于文件系统翻车了,相关日志内容缺失,所以本文我仅凭记忆描述事件,就不提供准确的日志输出了。 事件经过 出事的是 archlinuxcn 的编译机。那天中午时分我就收到了 Grafana 给我发送的莫名其妙的报警邮件,称某个监控项无数据了。我去面板上瞅了半天,明明有数据的啊。不解,但是忙别的事情去了,也没有细究。...

@lina archlinuxcn community encountered this about a year ago. One of the leaders lilydjwg blogged this event and followups on both her blog at blog.lilydjwg.me and the archcn(maybe) wiki. The symptom was full disk with too large commit that didn't fit in global reserve.
@lina Typical "btw i use arch" style?
@frd @Merritt @kde Geometric Weather😭
@lina @renedudfield Sounds like FreeBSD's linux compat layer...