Drop IO priority to idle while reading blocks for peer requests and startup verification #9245

luke-jr · 2016-11-30T08:29:55Z

No description provided.

src/util.h

laanwj · 2016-11-30T08:41:53Z

Concept ACK. Though not very happy to introduce platform-specific voodoo - we only just got rid of thread priority manipulation. But it may be worth the hassle, I don't know.

Can we quantify whether this works or not somehow?

gmaxwell · 2016-11-30T08:50:50Z

This will also delay other processing, in particular block relay-- at least until the handling is made more concurrent-- no? Not a reason to not do it, but maybe a reason to not do it by default for everyone.

I second the need to quantify this-- I could imagine it making for a big usability improvement. ... or not mattering at all. If the former, I want it... if the latter...

luke-jr · 2016-11-30T09:32:53Z

Whenever I restart my node lately, I find myself eventually manually ioniceing the entire process as it slows down other things monitoring it in iotop. I can't be sure it's sending out old blocks, but I can't imagine what else it'd be spending so much time reading... :/

Added Mac and Windows support for completeness.

ryanofsky

ACK 6430b9232048666996d5d4c3ed154907e3daff67 (after adding missing #includes)

ryanofsky · 2016-11-30T15:33:18Z

src/utilioprio.h

+#ifdef WIN32
+bool ioprio_set_file_idle(FILE *);
+#else
+#define ioprio_set_file_idle(f)  (false)


Maybe change this to ((void)false) to prevent a compiler warning:

main.cpp:1673:12: warning: statement has no effect [-Wunused-value] ioprio_set_file_idle(filein.Get());

ryanofsky · 2016-11-30T15:33:18Z

src/utilioprio.h

+#define IOPRIO_IDLER(actually_idle)  ioprio_idler ioprio_idler_(actually_idle)
+
+#else
+#define ioprio_get() (-1)


It doesn't seem like it should be necessary to declare these if the ioprio_idler class isn't around to call them.

We can simplify some of the other stuff (move the logic into the class itself) if low-level access is undesired, but for now it's too early to know if these won't be needed IMO.

ryanofsky · 2016-11-30T15:33:18Z

src/utilioprio.cpp

+// Distributed under the MIT software license, see the accompanying
+// file COPYING or http://www.opensource.org/licenses/mit-license.php.
+
+#ifdef HAVE_IOPRIO_SYSCALL


Needs #include "config/bitcoin-config.h" to prevent link errors.

ryanofsky · 2016-11-30T15:33:18Z

src/utilioprio.h

+// Distributed under the MIT software license, see the accompanying
+// file COPYING or http://www.opensource.org/licenses/mit-license.php.
+
+#ifndef BITCOIN_UTIL_IOPRIO_H


Should add #include "config/bitcoin-config.h"

fanquake · 2016-12-01T00:18:20Z

Travis failure:

'../../src/'`utilioprio.cpp
In file included from ../../src/utilioprio.cpp:9:0:
../../src/utilioprio.h: In destructor ‘ioprio_idler::~ioprio_idler()’:
../../src/utilioprio.h:42:51: error: ‘LogPrintf’ was not declared in this scope
             LogPrintf("failed to restore ioprio\n");

luke-jr · 2016-12-01T08:51:20Z

Looks like to make the Windows part work, we need to bump _WIN32_WINNT to 0x0600 which means it will only run on Vista or newer. AFAIK this is okay(?), but I'm going to leave it for a separate PR...

rebroad · 2016-12-19T04:35:40Z

I like this (concept ACK) although I wonder what the impact is on the p2p network as a whole if everyone ran this.

martinschwarz · 2017-03-11T19:57:41Z

Looks like to make the Windows part work, we need to bump _WIN32_WINNT to 0x0600 which means it will only run on Vista or newer.

There are win32 and win64 builds. Can't this just be enabled on the win64 build only?

laanwj · 2017-03-13T13:55:12Z

Looks like to make the Windows part work, we need to bump _WIN32_WINNT to 0x0600 which means it will only run on Vista or newer. AFAIK this is okay(?), but I'm going to leave it for a separate PR...

Isn't Vista the version after Windows XP? As we dropped support for Windows XP in 0.13, it seems that requiring Vista for 0.15 is fine.

There are win32 and win64 builds. Can't this just be enabled on the win64 build only?

Could be done, but it'd be confusing to couple those. The low-end systems running 32-bit versions would probably need this more.

luke-jr · 2017-08-21T05:06:48Z

Rebased...

TheBlueMatt · 2017-08-21T15:31:16Z

Hmm, I dont think this is really the best idea as long as our message processing is still single-threaded. Really we need to refactor stuff so that block reading is async and the network processing can continue for other peers while we're serving blocks for peers in IBD, otherwise we may block receiving a new block longer than required.

luke-jr · 2017-08-21T16:04:52Z

That's somewhat independent from this issue. If users need to shut off their node to use their computer, the delay for processing a new block will be even longer.

ryanofsky · 2017-10-12T17:47:24Z

@TheBlueMatt @luke-jr, maybe a compromise would be to make this behavior configurable, and perhaps to default to dropping priority if user is running bitcoin-qt on a desktop.

TheBlueMatt · 2017-11-10T21:54:10Z

Another approach which might be simpler would be to have the validation.h-exposed versions of ReadBlockFromDisk drop io priority so that net_processing will use low priority when answering remote-node queries but connecting blocks will not. With 0.15 I/O when doing initial sync is somewhat better, so this may also be less of an issue now unless the user is running with -peerbloomfilters.

luke-jr · 2017-11-11T06:28:30Z

@TheBlueMatt That's exactly what this already does... priority is only dropped when serving peers, not when connecting blocks.

TheBlueMatt · 2017-11-11T22:57:20Z

@luke-jr I was referring to the possibility of not exposing a priority flag in validation.h's API - that seems a bit overkill IMO, as evidenced by the fact that there are now two ReadBlockFromDisk calls in net_processing which dont get the low-priority flag :p. Though that would also result in RPC ReadBlockFromDisk calls getting de-prioritized.

More importantly, I'm curious how much we need this anymore - it seems most of the complaints about I/O usage were primarily due to 0.13.1 preferential peering...On systems where your I/O is severely limited, I both don't know how much this will help (in my experience Linux' ionice is mostly worthless when it comes to desktop latency) and don't know if its not better to direct people towards maxuploadtarget or peerbloomfilters so as to avoid simply slowing down your peers because your I/O is too slow.

luke-jr · 2017-11-11T23:23:57Z

Before writing this, I generally ionice'd the entire bitcoind process to maintain system usability.

TheBlueMatt · 2017-11-16T20:59:37Z

Concept ACK. You need to mark the other ReadBlockFromDisks in net_processing low-priority as well.

sipa · 2018-03-06T17:11:05Z

Concept ACK, but needs rebase.

kristapsk · 2022-07-01T13:33:32Z

Needs rebase.

davidgumberg · 2022-09-06T20:05:11Z

@luke-jr

I have rebased this here.

Feel free to take this.

glozow · 2022-09-26T10:01:56Z

Closing as this has needed rebase for more than 1 year. Feel free to reopen if you get a chance to work on this again in the future, thanks!

…tartup verification

luke-jr · 2022-09-26T13:36:16Z

Rebased

Xekyo · 2022-10-12T17:50:06Z

Could you please add a description to the first comment?

MarcoFalke · 2022-10-19T07:51:20Z

From CI:

./.libs/libbitcoinkernel.so: undefined reference to `ioprio_set_idle()'

DrahtBot · 2022-10-19T09:33:00Z

🐙 This pull request conflicts with the target branch and needs rebase.

_{Want to unsubscribe from rebase notifications on this pull request? Just convert this pull request to a "draft".}

fanquake · 2022-12-06T10:40:07Z

Moving this to draft for now. Since re-opening (after being closed due to needing rebase for > a year), the request for a PR description hasn't been fullfilled, it's been failing to compile in the kernel CI, failing the tidy job, and has now been needing rebase again for nearly 2 months.

Feel free to undraft when all 4 issues have been addressed.

DrahtBot · 2023-03-06T01:38:16Z

There hasn't been much activity lately and the patch still needs rebase. What is the status here?

Is it still relevant? ➡️ Please solve the conflicts to make it ready for review and to ensure the CI passes.
Is it no longer relevant? ➡️ Please close.
Did the author lose interest or time to work on this? ➡️ Please close it and mark it 'Up for grabs' with the label, so that it can be picked up in the future.

laanwj reviewed Nov 30, 2016

View changes

src/util.h Outdated Show resolved Hide resolved

laanwj added the Block storage label Nov 30, 2016

luke-jr force-pushed the ionice branch from b9303aa to 6430b92 Compare November 30, 2016 09:33

ryanofsky approved these changes Nov 30, 2016

View changes

luke-jr force-pushed the ionice branch 2 times, most recently from 4060358 to 3b9eeae Compare December 2, 2016 04:08

luke-jr force-pushed the ionice branch from 3b9eeae to 4073580 Compare February 4, 2017 21:29

This was referenced Mar 25, 2017

0.14.0 is using up 100% Disk I/O #9968

Closed

Disk IO > 90% on when running 0.13.1 on Ubuntu #9051

Closed

luke-jr force-pushed the ionice branch from 4073580 to 1a9b7b1 Compare August 21, 2017 05:05

luke-jr force-pushed the ionice branch from 1a9b7b1 to e673bc7 Compare August 21, 2017 21:41

luke-jr force-pushed the ionice branch from e673bc7 to 9e161f0 Compare March 6, 2018 19:20

DrahtBot mentioned this pull request Sep 12, 2021

allow -loadblock blocks to be unsorted #20331

Closed

DrahtBot added the Needs rebase label Sep 16, 2021

glozow closed this Sep 26, 2022

glozow reopened this Sep 26, 2022

luke-jr force-pushed the ionice branch from 6de915d to 26996ba Compare September 26, 2022 13:21

luke-jr added 3 commits September 26, 2022 13:35

Drop IO priority to idle while reading blocks for peer requests and s…

91e407b

…tartup verification

util/ioprio: Add Mac support using iopolicy functions

7e442a3

LoadExternalBlockFile: Set low I/O priority

b4647b2

luke-jr force-pushed the ionice branch from 26996ba to b4647b2 Compare September 26, 2022 13:36

DrahtBot removed the Needs rebase label Sep 26, 2022

This was referenced Sep 26, 2022

refactor: Replace m_params with chainman.GetParams() #25830

Merged

kernel: move RunCommandParseJSON to its own file #26196

Merged

MarcoFalke mentioned this pull request Oct 17, 2022

net: don't lock cs_main while reading blocks in net processing #26326

Open

DrahtBot mentioned this pull request Oct 17, 2022

use shared mutex to guard against block files being removed before read #26316

Closed

DrahtBot added the Needs rebase label Oct 19, 2022

fanquake marked this pull request as draft December 6, 2022 10:40

lyricidal mentioned this pull request Mar 21, 2023

[Upstream] Move Win32 defines to configure.ac to ensure they are globally defined PRCYCoin/PRCYCoin#388

Merged

Drop IO priority to idle while reading blocks for peer requests and startup verification #9245

Drop IO priority to idle while reading blocks for peer requests and startup verification #9245

luke-jr commented Nov 30, 2016

laanwj commented Nov 30, 2016 •

edited

gmaxwell commented Nov 30, 2016

luke-jr commented Nov 30, 2016

ryanofsky left a comment •

edited

ryanofsky Nov 30, 2016

ryanofsky Nov 30, 2016

luke-jr Nov 30, 2016

ryanofsky Nov 30, 2016

ryanofsky Nov 30, 2016

fanquake commented Dec 1, 2016

luke-jr commented Dec 1, 2016

rebroad commented Dec 19, 2016

martinschwarz commented Mar 11, 2017

laanwj commented Mar 13, 2017

luke-jr commented Aug 21, 2017

TheBlueMatt commented Aug 21, 2017

luke-jr commented Aug 21, 2017

ryanofsky commented Oct 12, 2017

TheBlueMatt commented Nov 10, 2017

luke-jr commented Nov 11, 2017

TheBlueMatt commented Nov 11, 2017

luke-jr commented Nov 11, 2017

TheBlueMatt commented Nov 16, 2017

sipa commented Mar 6, 2018

kristapsk commented Jul 1, 2022

davidgumberg commented Sep 6, 2022

glozow commented Sep 26, 2022

luke-jr commented Sep 26, 2022

Xekyo commented Oct 12, 2022

MarcoFalke commented Oct 19, 2022 •

edited

DrahtBot commented Oct 19, 2022

fanquake commented Dec 6, 2022

DrahtBot commented Mar 6, 2023

Drop IO priority to idle while reading blocks for peer requests and startup verification #9245

Are you sure you want to change the base?

Drop IO priority to idle while reading blocks for peer requests and startup verification #9245

Conversation

luke-jr commented Nov 30, 2016

laanwj commented Nov 30, 2016 • edited

gmaxwell commented Nov 30, 2016

luke-jr commented Nov 30, 2016

ryanofsky left a comment • edited

Choose a reason for hiding this comment

ryanofsky Nov 30, 2016

Choose a reason for hiding this comment

ryanofsky Nov 30, 2016

Choose a reason for hiding this comment

luke-jr Nov 30, 2016

Choose a reason for hiding this comment

ryanofsky Nov 30, 2016

Choose a reason for hiding this comment

ryanofsky Nov 30, 2016

Choose a reason for hiding this comment

fanquake commented Dec 1, 2016

luke-jr commented Dec 1, 2016

rebroad commented Dec 19, 2016

martinschwarz commented Mar 11, 2017

laanwj commented Mar 13, 2017

luke-jr commented Aug 21, 2017

TheBlueMatt commented Aug 21, 2017

luke-jr commented Aug 21, 2017

ryanofsky commented Oct 12, 2017

TheBlueMatt commented Nov 10, 2017

luke-jr commented Nov 11, 2017

TheBlueMatt commented Nov 11, 2017

luke-jr commented Nov 11, 2017

TheBlueMatt commented Nov 16, 2017

sipa commented Mar 6, 2018

kristapsk commented Jul 1, 2022

davidgumberg commented Sep 6, 2022

glozow commented Sep 26, 2022

luke-jr commented Sep 26, 2022

Xekyo commented Oct 12, 2022

MarcoFalke commented Oct 19, 2022 • edited

DrahtBot commented Oct 19, 2022

fanquake commented Dec 6, 2022

DrahtBot commented Mar 6, 2023

laanwj commented Nov 30, 2016 •

edited

ryanofsky left a comment •

edited

MarcoFalke commented Oct 19, 2022 •

edited