remove start stop restart daemonize #921
Labels
No Label
Benchmarking and Performance
HTTP Storage Protocol
Nevow Removal
Python 3 Porting
not-for-merge
No Milestone
No project
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies
No dependencies set.
Reference: tahoe-lafs/tahoe-lafs#921
Loading…
Reference in New Issue
No description provided.
Delete Branch "3550.remove-start-stop-restart-daemonize"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
https://tahoe-lafs.org/trac/tahoe-lafs/ticket/3550
https://tahoe-lafs.org/trac/tahoe-lafs/ticket/3523
https://tahoe-lafs.org/trac/tahoe-lafs/ticket/3524
Codecov Report
92% <ø> (ø)
74% <ø> (-1%)
63% <62%> (-37%)
65% <0%> (-19%)
95% <0%> (-<1%)
96% <0%> (+<1%)
95% <0%> (+1%)
Continue to review full report at Codecov.
@ -32,3 +23,1 @@
tahoe put -d DIR FILE testgrid:recentdir/recent.MD5SUM
echo "" | tahoe put -d DIR --mutable testgrid:log
echo "" | tahoe put -d DIR --mutable testgrid:recentlog
tahoe create-client --introducer=INTRODUCER_FURL DIR
Still need to do something about this...
There's one outstanding issue (check_grid.py) but I think the solution will be small and localized. The rest of this should be ready for a review.
Looks good to me!
Since this PR contains a fairly large set of changes, it might be good to have a second set of eyes to go over this?
@ -135,3 +133,4 @@
"``gatherer.tac``" file should be modified to add classifier functions.
The incident gatherer writes incident names (which are simply the relative
pathname of the ``incident-\*.flog.bz2`` file) into ``classified/CATEGORY``.
I am unsure about
flogtool create-incident-gatherer WORKDIR
because:Perhaps working examples would be better?
In the case of
flogtool create-gatherer
too, I believe a working example would be better.After start/stop/daemonize has been removed, will lines 219-261 still be up-to-date?
e59a922b27/src/allmydata/scripts/run_common.py (L219-L261)
Curious comment here. There's no special handling for cygwin in this module, so I guess this must have been stale since
629185d98
.We don't know anything about Cygwin these days, do we? I guess we will burn that bridge if/when we ever come to it again. :-)
There's also
DaemonizeTahoeNodePlugin
andDaemonizeTheRealService
that can probably go.Seems that
--nodaemon
argument totahoe run
command will not have any effect any more? There's handling for--nodaemon
in a bunch of places:51e50671e5/src/allmydata/test/cli/test_cli.py (L1285)
51e50671e5/src/allmydata/test/cli/test_cli.py (L1305-L1308)
51e50671e5/src/allmydata/scripts/tahoe_run.py (L14-L15)
Perhaps
scrips.tahoe_run
andscripts.run_common
can merge at this point?@ -135,3 +133,4 @@
"``gatherer.tac``" file should be modified to add classifier functions.
The incident gatherer writes incident names (which are simply the relative
pathname of the ``incident-\*.flog.bz2`` file) into ``classified/CATEGORY``.
Working examples are great. I guess this is a pre-existing and unrelated defect in the docs, though, so it shouldn't strictly be a blocker for merging this PR. I can file a ticket for this and the other error in this doc ... but my preferred outcome would probably be to escape Foolscap entirely and no longer have an "incident gatherer" or a "[log] gatherer" (and why is Tahoe-LAFS documenting the Foolscap
flogtool
anyway instead of referencing Foolscap-maintained docs anyway)?Also there's the larger question of CI for examples which we should have sometime.
So, gonna leave this broken but I filed a ticket - https://tahoe-lafs.org/trac/tahoe-lafs/ticket/3556
As far as I can tell, `DaemonizeTahoeNode", "DaemonizeTahoeNodePlugin", and "DaemonizeTheRealService" are somewhat misnamed as none of them actually do any daemonization - they're just the code that runs after daemonization is done.
But, yea, there is some more cleanup that could happen here, and maybe some more dead code that could be deleted. I'll see what else I can rip out.
Yea I haven't cared about Cygwin for a long time.
Turns out there wasn't really much.
I was looking at
check_grid.py
. This paragraph in the commentary is no longer correct, now that start/stop is gone:34cd1efaa4/src/allmydata/test/check_grid.py (L19-L21)
Setup instructions could be a little more helpful here:
34cd1efaa4/src/allmydata/test/check_grid.py (L25-L27)
In order to save some time with fiddling, I think lines 25-26 could be re-written as:
I found the error message when running
daemonize tahoe
confusing:It is unclear what is printing that message -- daemoinize or tahoe? (It is daemonize.) It should probably be:
Absolute paths to both
tahoe
andDIR
are needed, or things will be confusing. My copy of daemonize was silent when tahoe failed to run becauseDIR
was not an absolute path.Once all the setup was done
python check_grid.py DIR $(which tahoe)
reported that it's checked things it was supposed to check, without spewing a bunch of errors. I guess there could be a version ofcheck_grid.py
that can be nifty, with some work...