reload failure on upgrade to urbit-os-v2.119

[philipcmonk]

I received this error while receiving the OTA this morning. Looks like a failed scry, not sure if graph-store isn't ready to be scried from, or if it's scrying for something that doesn't exist? I'm guessing the former. That notebook exists and contains two posts. I have deleted posts from it and updated the ones that exist, if it makes a difference

...
kiln: bump %landscape to [%zuse 419]
kiln: %landscape from [~marzod %landscape] already at [%zuse 419]
kiln: bump %lfs-client to [%zuse 419]
kiln: %lfs-client from [~nilsud-walwyd-tabnus-fondeg %lfs-client] already at [%zuse 419]
gall: %hark-system-hook dozing, got %fact
gall: %docket dozing, got %fact
gall: reviving %gora
>>  [%gora %reload %sail]
eyre: replacing existing binding at /apps/gora
gall: reviving %face
gall: reviving %zone
gall: reviving %docs
gall: reviving %pals
gall: reviving %verse
gall: reviving %cliff
gall: reviving %chess
gall: reviving %docket
gall: reviving %treaty
gall: reviving %hark-store
gall: reviving %hark-system-hook
gall: reviving %settings-store
gall: reviving %naive-nmi
gall: reviving %naive-market
gall: reviving %mailer
gall: reviving %pipe
peek bad result
/sys/vane/gall/hoon:<[1.199 7].[1.200 42]>
/sys/vane/gall/hoon:<[1.260 35].[1.260 48]>
/sys/vane/gall/hoon:<[1.260 27].[1.260 48]>
/sys/vane/gall/hoon:<[1.260 26].[1.260 60]>
crud: %drip-fail event failed
drip failed
[%poke %wake]
%load
wire=/sys/cor/pipe/~wicdev-wisryt/studio/~2022.1.18..17.58.15..5f27

...

/sys/vane/gall/hoon:<[1.337 7].[1.341 36]>
/sys/vane/gall/hoon:<[1.339 7].[1.341 36]>
/sys/vane/gall/hoon:<[1.341 7].[1.341 36]>
%load-failed
/ gx
  ~wicdev-wisryt
  graph-store
  ~2022.1.18..17.58.15..5f27
  graph
  ~wicdev-wisryt
  wicdev-wisryt-s-weblog-3784
  node
  children
  kith
  ~
  ~
  noun
/lib/graph/hoon:<[14 3].[20 5]>
/app/pipe/hoon:<[794 5].[795 70]>
/app/pipe/hoon:<[793 3].[820 16]>
/app/pipe/hoon:<[788 3].[820 16]>
/app/pipe/hoon:<[836 7].[836 43]>
/app/pipe/hoon:<[834 3].[843 5]>
/app/pipe/hoon:<[833 3].[843 5]>
/app/pipe/hoon:<[832 3].[843 5]>
/app/pipe/hoon:<[884 33].[884 60]>
/app/pipe/hoon:<[884 18].[884 61]>
/app/pipe/hoon:<[884 3].[886 46]>
/app/pipe/hoon:<[883 3].[886 46]>
/app/pipe/hoon:<[881 3].[886 46]>
/app/pipe/hoon:<[880 3].[886 46]>
/app/pipe/hoon:<[52 7].[52 35]>
/app/pipe/hoon:<[51 5].[53 40]>
/app/pipe/hoon:<[49 5].[53 40]>
/app/pipe/hoon:<[48 5].[53 40]>
/app/pipe/hoon:<[47 5].[53 40]>
/app/pipe/hoon:<[46 5].[53 40]>
/app/pipe/hoon:<[45 5].[53 40]>
/app/pipe/hoon:<[43 3].[58 5]>
/app/pipe/hoon:<[42 3].[58 5]>
/app/pipe/hoon:<[41 3].[58 5]>
/app/pipe/hoon:<[40 3].[139 5]>
/app/pipe/hoon:<[39 3].[139 5]>
/sys/vane/gall/hoon:<[1.496 13].[1.496 49]>

Workaround: I suspended studio and the OTA went through fine.

Let me know if I can provide more information.

[philipcmonk]

In fact, after suspending studio and making the OTA go through again, I was able to revive studio with no problems. This seems like strong evidence that the issue is %pipe getting revived before %graph-store.

[mrdomino]

weirdly I did not have this problem on either of two ships running %studio.

[ixv]

I don't understand why studio's on-load is being called at all

[philipcmonk]

Zuse changed, so all apps have to be reloaded.

[ixv]

Oh I see. I suppose the thing to do here is check if graph-store has loaded, then either continue with the scry or queue up a move to do it later. I think there's a gall scry I can run to check this, could you remind me what the different gall scry cares are?

[ixv]

Nevermind, got it, looks like its %ge

[philipcmonk]

Yes I think %ge or %gu should work

[ixv]

Fixed