Introduce zfs rewrite subcommand #17246
Conversation
github-actions
bot
added
the
Status: Work in Progress
|
I've tried to find some kernel APIs to wire this to, but found that plenty of Linux file systems each implement their own IOCTL's for similar purposes. I did the same, except the IOCTL number I took almost arbitrary, since ZFS seems quite rough in this area. I am open to any better ideas before this is committed. |
|
This looks amazing! Not having to sift through half a dozen shell scripts every time this comes up to see what currently handles the most edge cases correctly is very much appreciated. Especially with RaidZ expansion, being able to direct users to run a built-in command instead of debating what script to send them to would be very nice. Also being able to reliably rewrite a live dataset while it's in use without having to worry about skipped files or mtime conflicts would make the whole process much less of a hassle. With the only thing to really worry about being snapshots/space usage this seems as close to perfect as reasonably possible (without diving deep into internals and messing with snapshot immutability). Bravo! |
amotin
added
the
Status: Design Review Needed
|
thank you. Fixes one of the biggest problems with ZFS. Is there a way to suspend the process? It might be nice to have it run only during off hours. |
It does one file at a time, and should be killable in between. Signal handling within one huge file can probably be added. Though the question of the process restart is on the user. I didn't plan to go that deep into the area within this PR. |
|
I couldn't find documentation in the files changed, so I have to guess how it actually works. Is it a file at a time? I guess you could feed it with a "find" command. For a system with a billion files, do you have a sense how long this is gong to take? We can do scrubs in a day or two, but rsync is impractically slow. If this is happening at the file system level, that migth be the case here as well. |
This will likely be a good use case for GNU Parallel. |
It can take a directory as an argument and there are some recursive functions and iterators in the code so piping find into it should not be necessary. That avoids some userspace file handling overhead, but it still has to go through the contents of each directory one file at a time. I also don't see any parallel execution or threading (though I'm not too familiar with ZFS internals, maybe some of the primitives used here run asynchronously?). Whether doing parallelism in userspace by just calling it for many files/directories at once or not it should have the required locking to just run in the background and be significantly more elegant than the CP + mtime (or potentially userspace hash) check to make sure files didn't change during the copy process avoiding one of the potential pitfalls of existing solutions. |
|
I haven't benchmarked it deep yet, but unless the files are tiny, I don't expect there is a major need for parallelism. The code in kernel should handle up to 16MB at a time, plus allows ZFS to do read-ahead and write-back on top of that, so there will be quite a lot in the pipeline to saturate the disks and/or the system, especially if there is some compression/checksuming/encryption. And without need to copy data to/from user-space, the only thread will not be doing too much, I think mostly a decompression from ARC. Bunch of small files on a wide HDD pool I suspect may indeed suffer from read latency, but that in user-space we can optimize/parallelize all day long. |
|
I gave this a quick test. It's very fast and does exactly what it says 👍 I can already see people writing scripts that go though every dataset, setting the optimal compression, recordsize, etc, and zfs rewrite-ing them. |
|
Cool! Though the recordsize is one of things it can't change, since it would requite real byte-level copy, not just marking existing blocks dirty. I am not sure it can be done under the load in general. At least it would be much more complicated. |
|
Umm this is basically same as doing send | recv, isn't it? I mean, in a way, this is already possible to do without any changes, isn't it? Recv will even respect a lower recordsize, if I'm not mistaken - at least when receiving into a pool without large blocks support, it has to do that. I'm thinking whether we can do better, in the original sense of ZFS "better", meaning "automagic" - what do you think of using snapshots, send|recv, in a loop with ever decreasing delta size and then when the delta isn't decreasing anymore, we could swap those datasets and use (perhaps slightly modified) It'd be even cooler if it could coalesce smaller blocks into larger ones, but that potentially implies performance problems with write amplification, I would say if the app writes in smaler chunks that it gets onto disk in such smaller chunks, it's probably for the best to leave them that way. For any practical use-case I could think of though, I would definitely appreciate the ability to split the blocks of a dataset using smaller If there's a way how to make |
|
send recv has the huge downside of requiring 2x the space, even if you do the delta size thing since it has to send the entire dataset at least once and old data can't be deleted until the new dataset is complete.
Isn't this exactly what rewrite does? Change the options, run it and all the blocks are changed in the background. Without an application even seeing a change to the file. And unlike send recv it only needs a few MB of extra space. Edit: with the only real exception being record size, but recv also solves that only partially at best and it doesn't look like there's a reasonable way to work around that in a wholly transparent fashion. |
|
amotin
force-pushed
the
rewrite
branch
4 times, most recently
from
d23a371 to
c5f4413
Compare
|
Which release is this game changing enhancement likely to land in? |
|
@stuartthebruce So far it haven't landed even in master, so anybody who want to speed it up is welcome to test and comment. In general though, when completed, there is no reason why aside of 2.4.0 it can't be ported back to some 2.3.x of the time. |
Good to know there are no obvious blockers from including in a future 2.3.x. Once this hits master I will help by setting up a test system with 1/2PB of 10^9 small files to see if I can break it. Is there any reason to think the code will be sensitive to Linux vs FreeBSD? |
IOCTL interface of the kernels is obviously slightly different, requiring OS-specific shims, as with most of other VFS-related code. But seems like not a big problem, as Tony confirmed it works on Linux too from the first try. |
Since this introduces new IOCTL API, I'd appreciate some feedback before it hit master in case some desired functionality might require API changes aside of the |
OK, I will see if I can find some time this next week to stress test. |
|
@stuartthebruce To specify any times we'd at very least need #16853 to land first. |
|
Added some tests (not yet tested ;)) and opaque arg field into the IOCTLs structure to make it even more future-proof. |
amotin
removed
the
Status: Design Review Needed
amotin
force-pushed
the
rewrite
branch
7 times, most recently
from
1e50c20 to
12b0e14
Compare
@tonyhutter Added and passed. |
|
You'll want to add some input validation to the |
Values beyond the file size are not illegal there. Kernel will rewrite only what is actually there. Can just add a check for non-numeric value, if you prefer. |
We should error out if a user is trying to seek/rewrite pass the end of the file. Maybe they accidentally typed in the wrong offset? Also, I'm wondering if we should not allow |
I don't think we should, considering the code was planned to work under concurrent load. We should not fail if the file just got truncated. We achieved our goal by doing nothing.
It may be a weird combination, but again not illegal. I am actually verifying it in the test, just because I can. |
These have now finished after 90 hours without any obvious problems, [root@zfsarchive1 ~]# zfs list
NAME USED AVAIL REFER MOUNTPOINT
jbod17 16.8T 910T 136K /jbod17
jbod17/cal 534G 910T 534G /jbod17/cal
jbod17/dqr 4.06T 910T 3.97T /jbod17/dqr
jbod17/grb.exttrig 4.99T 910T 4.98T /jbod17/grb.exttrig
jbod17/idq 2.51T 910T 2.51T /jbod17/idq
jbod17/pe.o4 4.73T 910T 4.63T /jbod17/pe.o4
[root@zfsarchive1 ~]# parallel 'zfs set compression=off {} && time zfs rewrite -r /{}' ::: jbod17/dqr jbod17/grb.exttrig jbod17/idq jbod17/pe.o4
real 616m48.074s
user 0m18.245s
sys 52m49.744s
real 1188m48.153s
user 0m14.149s
sys 59m4.946s
real 1544m41.673s
user 0m34.218s
sys 77m23.340s
real 5414m12.459s
user 2m19.411s
sys 135m19.536s
[root@zfsarchive1 ~]# zfs list
NAME USED AVAIL REFER MOUNTPOINT
jbod17 39.6T 887T 136K /jbod17
jbod17/cal 534G 887T 534G /jbod17/cal
jbod17/dqr 8.63T 887T 4.69T /jbod17/dqr
jbod17/grb.exttrig 11.4T 887T 6.40T /jbod17/grb.exttrig
jbod17/idq 9.13T 887T 6.62T /jbod17/idq
jbod17/pe.o4 9.95T 887T 5.23T /jbod17/pe.o4
|
Ok, I still think this is weird, but am not going to let it hold up the review. Please document the behavior though: diff --git a/man/man8/zfs-rewrite.8 b/man/man8/zfs-rewrite.8
index 5fb360d90..86c75e06e 100644
--- a/man/man8/zfs-rewrite.8
+++ b/man/man8/zfs-rewrite.8
@@ -54,6 +54,13 @@ Print names of all successfully rewritten files.
.It Fl x
Don't cross file system mount points when recursing.
.El
+.Pp
+Note: If a
+.Fl l
+or
+.Fl o
+value would cause a rewrite to regions past the end of the file, then those
+regions are silently ignored, and no error is reported.
.
.Sh SEE ALSO
.Xr zfsprops 7 |
|
I was using this built against 2.3.2 using I don't know if documenting that possibility might be helpful to users. |
amotin
force-pushed
the
rewrite
branch
2 times, most recently
from
b31fbb6 to
ed5405b
Compare
|
@satmandu I've added few more notes about possible effects. |
cmd/zfs/zfs_main.c
Outdated
| if (stat(path, &st) < 0) { | ||
| ret = errno; | ||
| (void) fprintf(stderr, gettext("failed to stat %s: %s\n"), | ||
| path, strerror(errno)); | ||
| return (ret); | ||
| } | ||
|
|
||
| if (S_ISREG(st.st_mode)) { | ||
| ret = zfs_rewrite_file(path, verbose, args); | ||
| } else if (S_ISDIR(st.st_mode) && recurse) { | ||
| ret = zfs_rewrite_dir(path, verbose, xdev, st.st_dev, args, | ||
| dirs); | ||
| } |
There was a problem hiding this comment.
I messed with this a bit, and came up with:
| if (stat(path, &st) < 0) { | |
| ret = errno; | |
| (void) fprintf(stderr, gettext("failed to stat %s: %s\n"), | |
| path, strerror(errno)); | |
| return (ret); | |
| } | |
| if (S_ISREG(st.st_mode)) { | |
| ret = zfs_rewrite_file(path, verbose, args); | |
| } else if (S_ISDIR(st.st_mode) && recurse) { | |
| ret = zfs_rewrite_dir(path, verbose, xdev, st.st_dev, args, | |
| dirs); | |
| } | |
| if (lstat(path, &st) < 0) { | |
| ret = errno; | |
| (void) fprintf(stderr, gettext("failed to stat %s: %s\n"), | |
| path, strerror(errno)); | |
| return (ret); | |
| } | |
| if (S_ISREG(st.st_mode)) { | |
| ret = zfs_rewrite_file(path, verbose, args); | |
| } else if (S_ISDIR(st.st_mode) && recurse) { | |
| ret = zfs_rewrite_dir(path, verbose, xdev, st.st_dev, args, | |
| dirs); | |
| } else { | |
| ret = S_ISDIR(st.st_mode) ? EISDIR : EINVAL; | |
| const char *errstr = | |
| (ret == EISDIR) ? "is a directory" : "not a regular file"; | |
| (void) fprintf(stderr, gettext("can't rewrite %s: %s\n"), | |
| path, errstr); | |
| } |
That is:
- don't allow a rewrite through a symlink (matches directory mode, protects against cross-device, protects multiple links to same file)
- show an error if you try to rewrite something that isn't a file
- show an error if you try to write a directory without
-r
The symlink thing I think is important; at least, it surprised me that it worked at all (rewrote the target), and not being the same as directory mode bothered me. The other two are just niceties.
I did think about also limiting to things not on ZFS, because it is strange to me that a zfs command could even have an opinion about non-ZFS filesystems. But, there isn't really a good cross-platform way to find this out, and the ioctl() call will probably fail, so I'm kind of whatever about it.
There was a problem hiding this comment.
About lstat() is a good point. Thank you. Fixed.
About the reporting errors I am not sure. I wanted to make possible running it with * not worrying about directory content, and when recursing we do not report errors for non-regular files, so it is kind of consistent.
Limiting to ZFS would be good, but I don't know how. That was actually the first reason I added the -x flag, to at least allow correct use of it.
There was a problem hiding this comment.
Ahh, I hadn't thought about *, interesting.
The actual thing that confused me originally was zfs-rewrite somedir/, which outputs nothing and returns success. And then it seemed like you might want that for other file types too.
I guess if I had to argue, I'd want more error output, not less, because I can redirect errors away, and I can prefilter by file types if I want with find | xargs.
(Other programs tackle this with a "quiet" option or, less often, a "noisy" option, but I never remember those sort of things exist)
| } | ||
|
|
||
| free(fullname); | ||
| } |
There was a problem hiding this comment.
Should we break the loop for certain kinds of errors from rewrite, like ENOSPC or EDQUOT? Probably not I guess, it's hard to tell for sure that it's the right thing and ctrl-c is right there.
amotin
force-pushed
the
rewrite
branch
2 times, most recently
from
6c7d427 to
6e6f97f
Compare
This allows to rewrite content of specified file(s) as-is without modifications, but at a different location, compression, checksum, dedup, copies and other parameter values. It is faster than read plus write, since it does not require data copying to user-space. It is also faster for sync=always datasets, since without data modification it does not require ZIL writing. Also since it is protected by normal range range locks, it can be done under any other load. Also it does not affect file's modification time or other properties. Signed-off-by: Alexander Motin <[email protected]> Sponsored by: iXsystems, Inc.
Motivation and Context
For years users were asking for an ability to re-balance pool after vdev addition, de-fragment randomly written files, change some properties for already written files, etc. The closest option would be to either copy and rename a file or send/receive/rename the dataset. Unfortunately all of those options have some downsides.
Description
This change introduces new
zfs rewritesubcommand, that allows to rewrite content of specified file(s) as-is without modifications, but at a different location, compression, checksum, dedup, copies and other parameter values. It is faster than read plus write, since it does not require data copying to user-space. It is also faster for sync=always datasets, since without data modification it does not require ZIL writing. Also since it is protected by normal range range locks, it can be done under any other load. Also it does not affect file's modification time or other properties.How Has This Been Tested?
Manually tested it on FreeBSD. Linux-specific code is not yet tested.
Types of changes
Checklist:
Signed-off-by.