Issue 176727 in chromium: Heavy I/O generated by chrome during inactivity - what happen/some craches?

[chro...@googlecode.com]

Status: Unconfirmed
Owner: ----
Labels: Pri-2 Via-Wizard Type-Bug OS-Windows

New issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated by chrome
during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

UserAgent: Mozilla/5.0 (Windows NT 5.1) AppleWebKit/537.17 (KHTML, like
Gecko) Chrome/24.0.1312.57 Safari/537.17

Steps to reproduce the problem:
1. Just start chrome (but with some history)?
2.
3.

What is the expected behavior?
No I/O during inactivity since application should no use my cpu/hdd if not
need!

What went wrong?
Inactive chrome generate such I/O (task manager - it kills my disk)?
i/o read 8 200 000 000
i/o write 1 900 000 000

i/o pre minute is about (7% of cpu used constanly):
read + 200 000 000 + 50 0000 000

See what Chrome doing (what is it - process monitor sysinternals)?
chrome.exe 3056 0.0000000 159 914 159 914 159 914 60 792 60 792 834 834 834
n/a 0 n/a 0

Did this work before? N/A

Chrome version: 24.0.1312.57 Channel: stable
OS Version: 5.1 (Windows XP)

--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[chro...@googlecode.com]

Comment #1 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

Some log from process monitor (It generates thousand *.sst what it is? - I
put some for example):

426592 15:16:02,3656932 chrome.exe 0.0000246 3056 WriteFile C:\Documents
and Settings\Cezary Wagner\Local Settings\Application
Data\Google\Chrome\User Data\Default\Session Storage\LOG SUCCESS Offset:
727 031, Length: 59 Write "C:\Documents and Settings\Cezary Wagner\Local
Settings\Application Data\Google\Chrome\Application\chrome.exe" File
System 24.0.1312.57
426593 15:16:02,3662017 chrome.exe 0.0000374 3056 ReadFile C:\Documents and
Settings\Cezary Wagner\Local Settings\Application Data\Google\Chrome\User
Data\Default\Session Storage\091905.sst SUCCESS Offset: 0, Length: 10 538
Read "C:\Documents and Settings\Cezary Wagner\Local Settings\Application
Data\Google\Chrome\Application\chrome.exe" File System 24.0.1312.57
426594 15:16:02,3664129 chrome.exe 0.0000203 3056 ReadFile C:\Documents and
Settings\Cezary Wagner\Local Settings\Application Data\Google\Chrome\User
Data\Default\Session Storage\091917.sst SUCCESS Offset: 0, Length: 5 457
Read "C:\Documents and Settings\Cezary Wagner\Local Settings\Application
Data\Google\Chrome\Application\chrome.exe" File System 24.0.1312.57
426595 15:16:02,3665897 chrome.exe 0.0000581 3056 ReadFile C:\Documents and
Settings\Cezary Wagner\Local Settings\Application Data\Google\Chrome\User
Data\Default\Session Storage\091926.sst SUCCESS Offset: 0, Length: 10 205
Read "C:\Documents and Settings\Cezary Wagner\Local Settings\Application
Data\Google\Chrome\Application\chrome.exe" File System 24.0.1312.57
426596 15:16:02,3668274 chrome.exe 0.0000204 3056 ReadFile C:\Documents and
Settings\Cezary Wagner\Local Settings\Application Data\Google\Chrome\User
Data\Default\Session Storage\091929.sst SUCCESS Offset: 0, Length: 5 267
Read "C:\Documents and Settings\Cezary Wagner\Local Settings\Application
Data\Google\Chrome\Application\chrome.exe" File System 24.0.1312.57
426597 15:16:02,3670079 chrome.exe 0.0000260 3056 ReadFile C:\Documents and
Settings\Cezary Wagner\Local Settings\Application Data\Google\Chrome\User
Data\Default\Session Storage\091938.sst SUCCESS Offset: 0, Length: 5 256
Read "C:\Documents and Settings\Cezary Wagner\Local Settings\Application
Data\Google\Chrome\Application\chrome.exe" File System 24.0.1312.57
426598 15:16:02,3671990 chrome.exe 0.0000316 3056 ReadFile C:\Documents and
Settings\Cezary Wagner\Local Settings\Application Data\Google\Chrome\User
Data\Default\Session Storage\091944.sst SUCCESS Offset: 0, Length: 10 098
Read "C:\Documents and Settings\Cezary Wagner\Local Settings\Application
Data\Google\Chrome\Application\chrome.exe" File System 24.0.1312.57
426599 15:16:02,3674644 chrome.exe 0.0000240 3056 ReadFile C:\Documents and
Settings\Cezary Wagner\Local Settings\Application Data\Google\Chrome\User
Data\Default\Session Storage\091947.sst SUCCESS Offset: 0, Length: 10 824
Read "C:\Documents and Settings\Cezary Wagner\Local Settings\Application
Data\Google\Chrome\Application\chrome.exe" File System 24.0.1312.57
426600 15:16:02,3676490 chrome.exe 0.0000257 3056 ReadFile C:\Documents and
Settings\Cezary Wagner\Local Settings\Application Data\Google\Chrome\User
Data\Default\Session Storage\091968.sst

[chro...@googlecode.com]

Comment #2 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

The Chrome become number one in consuming i/o in system it need about 2
mins to beat all other processes - I am programmer and see that there is
serious error.

http://screencast.com/t/WlGQAKhe0jIn
http://screencast.com/t/fL2mQsqhm7t5

[chro...@googlecode.com]

Comment #3 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

I found some workaround:
1. go C:\Documents and Settings\Cezary Wagner\Local Settings\Application

Data\Google\Chrome\User Data\Default\Session Storage\

2. Delete all (why it works and what I delete I do not know but it stop
heavy i/o)???

Please explain what I have done if you are able?

[chro...@googlecode.com]

Comment #4 on issue 176727 by throwar...@gmail.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

Deleting the Session Storage folder worked for me as well.
Screenshot showing I/O chart:
https://dl.dropbox.com/u/6791352/ScreenShot168.png When I stopped chrome
and deleted the folder, the change was significant.

Attachments:
ScreenShot168.png 61.8 KB

[chro...@googlecode.com]

Comment #5 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

This bug could destroy disk drive and data with heavy traffic it should be
fixed since Chrome is currently efficient disk killer :)

[chro...@googlecode.com]

Comment #6 on issue 176727 by laboboL...@gmail.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

There have been many reports on Heavy I/O activity caused by chrome.
I have this problem as well.
Chrome is humping log files in 'Session Storage' folder. Hanging the
browser and making it slow.
On my laptop it is very noticeable.
SSD owners will get upset about this.

We need to bug a dev to get this fixed already. These sorts of problems
costs money.
Will google pay for my replacement hard drive?

[chro...@googlecode.com]

Comment #7 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

That is good question - if Google want damage hard disk and be devil should
pay money for it. Damages of all data can be very costly.

The best is that no one read this bug since it is not confirmed or
rejected :)

[chro...@googlecode.com]

Comment #8 on issue 176727 by nChase...@gmail.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

I have an older computer and chrome hogs all the disk io, freezing up the
rest of my programs. Thankfully the work around Cezary mentioned fixed
it... for now. I want to know why something as destructive as this bug
isn't being noticed. I'm going to bug the other chrome users I know and see
if this is an issue for them as well.

[chro...@googlecode.com]

Comment #13 on issue 176727 by ma...@chromium.org: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

When Chrome is idle, we do some session storage scavenging, deleting
unneeded data piece by piece, once per minute. However, this should only
cause small writes and doesn't sound like this should be behind this bug.

Can it be that LevelDB somehow generates a lot of activity in this case or
that we're using it in a pathological way?

[chro...@googlecode.com]

Updates:
Cc: dgr...@chromium.org
Labels: -Cr-Blink-Storage-IndexedDB

Comment #14 on issue 176727 by dgr...@chromium.org: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

I couldn't get a sense of the scale on the graph in comment 4, is that over
seconds, minutes, hours?

First, the log from comment 1 indicates chrome 24. This could be related to
the missing sst problem that was fixed in 25/26. There were reports of
affected users having their leveldb directories "blow up", where thousands
of sst files were created.

Second, it's possible that leveldb is being used pathologically, especially
if the database is > ~100mb-1gb and the keys being deleted are >4mb. What
we'd need is for someone who is experiencing this problem to move their
Session Storage directory aside. If that fixes the issue, send us the
directory so that we can try to reproduce locally.

[chro...@googlecode.com]

Comment #15 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

I think that you not think since you it is impossible what you want :)

Why I have to keep invalid data on disk since FEBRUARY?

Why I have to keep invalid data on disk since FEBRUARY if it destroys my
disc and slow down my productivity?

Learn from this log what is destroying disc and slow down computer - I will
not help you reproduce - you should ask about it one day after I detect
crash not 3 months later.

426592 15:16:02,3656932 chrome.exe 0.0000246 3056 WriteFile C:\Documents

and Settings\Cezary Wagner\Local Settings\Application

Data\Google\Chrome\User Data\Default\Session Storage\LOG SUCCESS Offset:

727 031, Length: 59 Write "C:\Documents and Settings\Cezary Wagner\Local

Settings\Application Data\Google\Chrome\Application\chrome.exe" File
System 24.0.1312.57

426593 15:16:02,3662017 chrome.exe 0.0000374 3056 ReadFile C:\Documents and

Settings\Cezary Wagner\Local Settings\Application Data\Google\Chrome\User

Data\Default\Session Storage\091905.sst SUCCESS Offset: 0, Length: 10 538
Read "C:\Documents and Settings\Cezary Wagner\Local Settings\Application
Data\Google\Chrome\Application\chrome.exe" File System 24.0.1312.57

426594 15:16:02,3664129 chrome.exe 0.0000203 3056 ReadFile C:\Documents and

Settings\Cezary Wagner\Local Settings\Application Data\Google\Chrome\User

Data\Default\Session Storage\091917.sst SUCCESS Offset: 0, Length: 5 457
Read "C:\Documents and Settings\Cezary Wagner\Local Settings\Application
Data\Google\Chrome\Application\chrome.exe" File System 24.0.1312.57

426595 15:16:02,3665897 chrome.exe 0.0000581 3056 ReadFile C:\Documents and

Settings\Cezary Wagner\Local Settings\Application Data\Google\Chrome\User

Data\Default\Session Storage\091926.sst SUCCESS Offset: 0, Length: 10 205
Read "C:\Documents and Settings\Cezary Wagner\Local Settings\Application
Data\Google\Chrome\Application\chrome.exe" File System 24.0.1312.57

426596 15:16:02,3668274 chrome.exe 0.0000204 3056 ReadFile C:\Documents and

Settings\Cezary Wagner\Local Settings\Application Data\Google\Chrome\User

Data\Default\Session Storage\091929.sst SUCCESS Offset: 0, Length: 5 267
Read "C:\Documents and Settings\Cezary Wagner\Local Settings\Application
Data\Google\Chrome\Application\chrome.exe" File System 24.0.1312.57

426597 15:16:02,3670079 chrome.exe 0.0000260 3056 ReadFile C:\Documents and

Settings\Cezary Wagner\Local Settings\Application Data\Google\Chrome\User

Data\Default\Session Storage\091938.sst SUCCESS Offset: 0, Length: 5 256
Read "C:\Documents and Settings\Cezary Wagner\Local Settings\Application
Data\Google\Chrome\Application\chrome.exe" File System 24.0.1312.57

426598 15:16:02,3671990 chrome.exe 0.0000316 3056 ReadFile C:\Documents and

Settings\Cezary Wagner\Local Settings\Application Data\Google\Chrome\User

Data\Default\Session Storage\091944.sst SUCCESS Offset: 0, Length: 10 098
Read "C:\Documents and Settings\Cezary Wagner\Local Settings\Application
Data\Google\Chrome\Application\chrome.exe" File System 24.0.1312.57

426599 15:16:02,3674644 chrome.exe 0.0000240 3056 ReadFile C:\Documents and

Settings\Cezary Wagner\Local Settings\Application Data\Google\Chrome\User

Data\Default\Session Storage\091947.sst SUCCESS Offset: 0, Length: 10 824
Read "C:\Documents and Settings\Cezary Wagner\Local Settings\Application
Data\Google\Chrome\Application\chrome.exe" File System 24.0.1312.57

426600 15:16:02,3676490 chrome.exe 0.0000257 3056 ReadFile C:\Documents and

Settings\Cezary Wagner\Local Settings\Application Data\Google\Chrome\User

Data\Default\Session Storage\091968.sst

[chro...@googlecode.com]

Comment #16 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

I think that you do not think since you it is impossible what you want :)

Why I have to keep invalid data on disk since FEBRUARY?

Why I have to keep invalid data on disk since FEBRUARY if it destroys my

disc and slows down my productivity?

Learn from this log what is destroying disc and slow down computer - I will

not help you reproduce bug and not want to overheat my disc to the death -
you should ask about it one day after I detect crash not 3 months later. Do
ask about what is impossible since it lead to nothing

[chro...@googlecode.com]

Comment #17 on issue 176727 by sfinx.so...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

The same issue observed under Linux. The bug appears while switching from
21.x to 22.x. All the chrome versions >21.x are unusable for me - I do not
intend to wait minutes before the data appears.

[chro...@googlecode.com]

Comment #18 on issue 176727 by matthew....@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

I also have issues with high I/O activity which is causing problems. I'm a
software developer and my machine is sometimes unusable. The problem is I
have come to rely on Chrome with LastPass and a couple of other extensions.
I have got in the habit of not having too many extensions installed as I
know it's a problem but Chrome's use of my laptop's resources is
unacceptably high. I'm a software developer and Visual Studio doesn't use
as much resource! My set up is the following:

500GB drive with 7200RPM
Quad processor
8GB RAM
Windows 8.1 (new installation)

Thanks in advance,

Matt

[chro...@googlecode.com]

Comment #19 on issue 176727 by daj...@vanadac.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

This bug persists into Chrome version 30.0.1599.101 m. Deleting the
Session Storage folder and tmp files does not resolve the problem.

I noticed this because my Windows 8.1 laptop was slow after installing
Chrome. The Resource Monitor reported that chrome.exe was using 100% of
the available Disk I/O capacity.

A side-effect of the constant disk access is reduced battery life and
reduced RST SSD cache performance.

[chro...@googlecode.com]

Comment #21 on issue 176727 by rab.u...@gmail.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

With Chrome v31.0 on Win7, the high and exhausting disk io load appears
again. Network action predictor and Session Storage are slowing down my SSD
due to 100% io WRITE load.
Btw. RAM load remains <0.5GB. Only disk io by WRITE load is affected.

[chro...@googlecode.com]

Comment #22 on issue 176727 by fritzmat...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

Hi, i have the same problem, but i think can mitigate this. First, the
Local Storage folder, holds all cookies from the sites and some extensions
config too. My solution was change the default cookie config to "block
sites to store information" and install some extensions to manage and debug
de Local Storage. Second, the local Storage has 2 file types, one of it are
sqlitle database data.

Regards

[chro...@googlecode.com]

Comment #24 on issue 176727 by raphael....@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

Hi,

It's been months since this problem is bugging me and making my whole
computer slow.
I found a fix that works :

Go there :
http://www.mozilla.org/fr/firefox/new/

Download, install ...

Seriously, Chrome as become crap recently, lot of annoying bugs, critical
issues not being addresses, consumes tremendous amount of Memory.
It's time to move on ...

[chro...@googlecode.com]

Comment #25 on issue 176727 by lampu...@gmail.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

It killed my 1TB hard disk last summer.
Now I just use it with ramdisk

[chro...@googlecode.com]

Comment #26 on issue 176727 by txsanyix: Heavy I/O generated by chrome

during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

Too much IO can't kill a HDD, but can kill SSD memory cells which are
writeable for limited times, but HDD doesn't have such limitations. HDDs
usually kill themselves, as those are mechanical devices.

[chro...@googlecode.com]

Comment #27 on issue 176727 by jmngo...@gmail.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

This is off-topic, but so was the previous post: mechanical devices have a
limited operating life, just like SSD. Workload kills HDDs. Chrome
certainly doesn't help when it reads 300MB upon starting...

[chro...@googlecode.com]

Comment #28 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

Looking comments I see that some people not understand problem at all.

Chrome can transform ABNORMALLY into tool generating VERY HEAVY i/o and
generate really HUGE i/o without reason that is problem - it is doing HEAVY
i/o in background without motivation.

Problem is not if Chrome generate i\o if you download large files or
generate heavy browsing that is NOT SUCH CASE.

Whatever recent version of Chrome is spies strokes on keyboard and click
without reason on harddisk generating some i/o - you can test it running
Sysinternals Process Monitor on Windows - Chrome without clicking in THIS
FORM will generate about 2xless i/o. I background THIS VERSION OF Chrome
generates also i/o without reason but not as high as described in problem.
Whatever at end of day it can be 2 GB of doing nothing usefull.

Option stopping background not working for me since chrome not dies after
exit and need taskkill to clean computer from it.

[chro...@googlecode.com]

Comment #29 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

BTW Firefox is not doing such background i/o - my Task Manager shows no i/o
if no activity for FireFox.

[chro...@googlecode.com]

Comment #30 on issue 176727 by rafaelka...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

Crazy i/o in the Session storage folder! Please fix this!

[chro...@googlecode.com]

Comment #31 on issue 176727 by a...@chromium.org: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

Issue 398966 has been merged into this issue.

[chro...@googlecode.com]

Comment #32 on issue 176727 by zhangyin...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

Chrome makes my hard disk keep zzzz. It's just doing nothing other than
replying this thread.

Fortunately I have redirected its data file to SATA instead of SSD.

F**K!!!

[chro...@googlecode.com]

Comment #34 on issue 176727 by ma...@chromium.org: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

Whoever is experiencing this problem and the target of the disk activity is
Session Storage - could you please zip your Session Storage folder and send
it to me?

Please note that it can contain personal data - basically the web pages you
visit can write anything they want there. Do *not* make it publicly
available or attach it to this bug. The best would be if you could
reproduce this with a clean profile, just by visiting some web pages.

I can only help in the cases where the abnormal disk activity targets the
Session Storage folder.

[chro...@googlecode.com]

Comment #35 on issue 176727 by ma...@chromium.org: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

(And in case my e-mail address is not visible for you, it's
ma...@chromium.org.)

[chro...@googlecode.com]

Comment #37 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
http://code.google.com/p/chromium/issues/detail?id=176727

I can not tolerate more Chrome and switch to Firefox.

I feel now and see in facts that Firefox is little more stable, generetes
almost no i/o, double reduce swamping and memory usage.

What is important esspecially in summer and for people Firefox generates
less hot and noise - you have to improve Chrome since what I see not it is
terrible in comparison to Firefox - maybe double resource consumption what
is quite visible after one day of usage.

[chro...@googlecode.com]

Comment #39 on issue 176727 by saurabhd...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?

https://code.google.com/p/chromium/issues/detail?id=176727

why google you are not fixing it,this is serius Issue. I am a big fan of
chrome but due to this problem I have to move from chrome to firefox.
Because of chrome my Original HD got chashed.

[chro...@googlecode.com]

Comment #40 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

Why Google is not able to fix feature destroying discs and charming work
performance???

Why in Google there is not good product managers - one year for such
horrible problem is very long?

[chro...@googlecode.com]

Comment #41 on issue 176727 by iceh...@gmail.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

Now I know why my last hard disk was damaged.

[chro...@googlecode.com]

Comment #42 on issue 176727 by nirva...@gmail.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I´m using Chrome on Windows 7 32bit I am experiencing high I/O Reads and
writes, I am going to switch to Firefox until this is not fixed. Its a
fresh installation of OS and Chrome that I installed week ago.

[chro...@googlecode.com]

Comment #43 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I am using Chrome if I have to since it degrades system performance and can
destroy hard discs.

[chro...@googlecode.com]

Comment #45 on issue 176727 by pullin.t...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

no answer, no fix. FO google

[chro...@googlecode.com]

Comment #46 on issue 176727 by Z3rga...@gmail.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

Just realized that my hard disk is about to be destroyed. Read speed is
about 15 MBytes/sec. Is it chrome? Do not know. But I am switching to
firefox now.

[chro...@googlecode.com]

Comment #47 on issue 176727 by assorted...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I don't know if it's the somewhat childish language people are using that's
concealing it, but this is a very serious bug. It is probably the most
serious non-security related bug that Chrome has, and it's wholly
unacceptable this bug has been "unconfirmed" for over a year now (on this
ticket, complaints and other tickets go back further). In many ways their
anger and frustration is justified by an apparent attitude of indifference
towards a bug that is seriously degrading performance of systems.

Leaving Chrome running in the background for 7 hours just generated over 66
million I/O write operations and 5 million I/O read operations resulting in
over 1.6 GBs of reads and 3 GBs of writes from the "main" chrome process
(not from the tabs that were opened). That averages out to about 0.045kb
per write operation.

Making millions of write operations at an average size of only 0.045kb is a
horrid waste of resources on multiple fronts. Writes like this will cause
excess activity on SSDs (and this activity will be well out of proportion
to that which simply writing large blocks totaling to 3GBs would incur).
Additionally, both PIO and DMA transfers of such small amounts will cause
degraded performance (albeit in slightly different manners). On some
systems even the noise involved can be a problem.

Until this is fixed, consider disabling Session Storage as a default, since
multiple people (including myself) have confirmed that these operations are
related to it. Extensions are not causing it, I've had this bug in the past
without extensions, and even now I've tested uninstalling them with no
difference. It's clear that the bug has existed for a long time, both
Chrome and Chromium have obviously shared it for this long, and it's time
for the project to face it.

[chro...@googlecode.com]

Updates:
Cc: dgr...@chromium.org

Comment #48 on issue 176727 by ma...@chromium.org: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

(LevelDB people, pls read the second part of this. dgrogan@, can you help
to redirect this to the relevant people?)

Let's do another attempt at getting this back on track:

Whoever is experiencing the problem and wants to help, please do the
following:

1) Move your Session Storage folder aside, zip it, make it available to
ma...@chromium.org with your favourite file sharing service (DropBox,
Google Drive). Please note that it might contain personal data - at the
very least it'll contain the domains of the web pages you've visited with
that profile. It won't contain any data from incognito.

2) Does the problem reoccur after you have deleted the Session Storage
folder? The best would be if you could reproduce this with a clean profile,
visiting only a small number of web pages where you don't need to log in -
this way your Session Storage folder won't contain personal data you don't
want to share.

-------------

LevelDB people (who's working on LevelDB atm, actually?): is there any way
to pull more logging out of LevelDB? The ideal would be a command line
switch to give to Chrome, and then the users could get the LevelDB logs and
send them to me. Does something like this exist? If not, we'd probably need
to add it. SG?

[chro...@googlecode.com]

Updates:
Cc: micha...@chromium.org

Comment #49 on issue 176727 by jsb...@chromium.org: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

cmumford@ for leveldb stuff in general
michaeln@ for session storage specifically

[chro...@googlecode.com]

Comment #51 on issue 176727 by ma...@chromium.org: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

Great! In that case, whoever is experiencing this problem, pls provide
the "LOG" from the Session Storage directory so we can have a look.

[chro...@googlecode.com]

Comment #52 on issue 176727 by charles....@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I think I am not affected by this.
Anyway, LOG and LOG.old attached.

Attachments:
LOG 640 bytes
LOG.old 624 bytes

[chro...@googlecode.com]

Comment #53 on issue 176727 by charles....@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

Ah, sorry. Forgot to write the system info.
Chrome 40.0.2194.2 dev (64-bit) on Arch Linux.

[chro...@googlecode.com]

Comment #54 on issue 176727 by ma...@chromium.org: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

Your logs look so clean (not much activity) that it indeed looks like
you're not affected by this :)

[chro...@googlecode.com]

Comment #55 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I try to send log if it happen - I am not using Chrome now since heavy i/o
is often - I am trying to close chrome as soon as possible since leaving
Chrome leads to system slow down with suspicious heavy i/o.

I observed at last two times per month that Chrome - when I am forced to
use Chrome by Google Business View Panorama Editor or want to have
Polish/Unicode letters in Google Drive documents - it not works in Firefox.

Firefox use less memory and works fine - it sometimes go into heavy CPU but
never in heavy i/o so my hardisk is safe.

[chro...@googlecode.com]

Comment #56 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I read logs and nothing special inside :)

Chrome become i/o zombie yesterday generating 30GB i/o before I killed
system poison.

Only 5 lines left.

C:\Users\Crezary Wagner\Local Settings\Application Data\Google\Chrome\User
Data\Default\Session Storage 14:24:17,11
$cat LOG
2014/10/31-09:10:36.414 6260 Recovering log #4501
2014/10/31-09:10:41.710 6260 Level-0 table #4503: started
2014/10/31-09:10:50.917 6260 Level-0 table #4503: 886 bytes OK
2014/10/31-09:11:18.870 6260 Delete type=0 #4501
2014/10/31-09:11:18.871 6260 Delete type=3 #4499

C:\Users\Crezary Wagner\Local Settings\Application Data\Google\Chrome\User
Data\Default\Session Storage 14:24:22,99
$cat LOG.old
2014/10/30-23:04:44.342 9076 Recovering log #4497
2014/10/30-23:04:44.342 9076 Level-0 table #4500: started
2014/10/30-23:04:44.417 9076 Level-0 table #4500: 399 bytes OK
2014/10/30-23:04:44.523 9076 Delete type=0 #4497
2014/10/30-23:04:44.523 9076 Delete type=3 #4495

How do you want solve problem if it not logged?

[chro...@googlecode.com]

Comment #57 on issue 176727 by ma...@chromium.org: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

Interesting. How do you know the heavy I/O is targeting this directory?

Do you have several user profiles in your Chrome?

[chro...@googlecode.com]

Comment #59 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I have today high i/o - chrome become crazy again - I open such tabs Google
Drive, Google Docs, Two other page, Google Business View Panoeditor (forced
to use Chrome - the reason I tun this).

I was deleted the most of extensions some months ago including gmail which
generates some additional i/o. I have only Changout all other is disabled.

I often open up to 12 pages and prefer no more than 7.

I am using procmon as you see in #1 comment - I try to capture it again but
try to learn #1 comment first. I am web apps programmer (15 years).

LOG.old before I kill I think - I was not focused but kill fast Chrome
after slow down - with heavy i/o

2014/11/04-19:00:17.080 42248 Recovering log #852
2014/11/04-19:00:17.080 42248 Level-0 table #854: started
2014/11/04-19:00:17.153 42248 Level-0 table #854: 6120 bytes OK
2014/11/04-19:00:17.260 42248 Delete type=0 #852
2014/11/04-19:00:17.265 42248 Delete type=3 #850
2014/11/04-19:49:47.706 22676 Compacting 3@0 + 1@1 files
2014/11/04-19:49:48.088 22676 Generated table #856: 377 keys, 1974641 bytes
2014/11/04-19:49:48.088 22676 Compacted 3@0 + 1@1 files => 1974641 bytes
2014/11/04-19:49:48.089 22676 compacted to: files[ 0 2 0 0 0 0 0 ]
2014/11/04-19:49:48.089 22676 Delete type=2 #846
2014/11/04-19:49:48.105 22676 Delete type=2 #848
2014/11/04-19:49:48.111 22676 Delete type=2 #851
2014/11/04-19:49:48.112 22676 Delete type=2 #854

Web history: I can not see on other computer - next bug but nothing special.

[chro...@googlecode.com]

Comment #62 on issue 176727 by laboboL...@gmail.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

There is a known issue for users using windows 7, afaik, the JumpListIcons
are going nuts for each tab closed, separately. Causing major hard-drive
activity creating icons. You can verify it by right clicking the chrome
icon in the taskbar, you will see them as generic icons during this
activity.

Chrome have already killed my EXPENSIVE Seagate hybrid hard-drive. Seagate
is garbage, btw!

[chro...@googlecode.com]

Comment #64 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I have good advice for all forced to use Chrome - install in scheduler
script killing chrome after some idle time or at specified time to protect
hard drives.

Never leave Chrome idle since it can damage disk in inpredictable maner - I
replaced already 2 disk in RAID 1+0 - good that I have RAID and no data is
lost only money for disk and work.

It is easy detect abnormal behaviour which is random and often - if system
slow down, air flow is high and the most i/o is done by Chrome in task
manager (or below antivir) - it is time to kill beast to save your time,
power and disk.

BTW it will be nice if I will receive two 500GB disc from Google as
refundation and reward for detection this serious bug.

TASKKILL [/S system [/U username [/P [password]]]]
{ [/FI filter] [/PID processid | /IM imagename] } [/T] [/F]

[chro...@googlecode.com]

Comment #65 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I can give you some clue that maybe Google Docs and Google Mail since it
generates some network activity but why there is i/o if no docs or mail
update.

If you focus on solving this bug I will focus too.

[chro...@googlecode.com]

Comment #66 on issue 176727 by ttoo4...@gmail.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I am so sick of Google Chrome crippling my computer.
These issues have been going on FOR YEARS!
FOR YEARS!!!!
THIS HAS BEEN GOING ON FOR YEARS!!!!
THIS HAS BEEN GOING ON FOR YEARS!!!!
THIS HAS BEEN GOING ON FOR YEARS!!!!
THIS HAS BEEN GOING ON FOR YEARS!!!!
That is pathetic.
It's the same nonsense over and over again from Google Techs blaming
everything else under the sun except Google Chrome.
I don't care why Chrome is making .tmp files.
I don't care why Chrome is making some .tmp files that can't be seen
through a folder.
Even if folder options are set to be able to see EVERYTHING - some of the
high I/O usage .tmp files are missing in the folders that Resource Manager
shows them in.
None the less missing or not the .tmp files keep sucking the life out of my
computer.
No respectable programmer would ever release such garbage to the public!
No respectable programmer would ever release such garbage to the public!
No respectable programmer would ever release such garbage to the public!
No respectable programmer would ever release such garbage to the public!
No respectable programmer would ever release such garbage to the public!
No respectable programmer would ever release such garbage to the public!

[chro...@googlecode.com]

Comment #67 on issue 176727 by assorted...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

What is happening with this issue? There seemed to be a flurry of activity
but the bug is still marked as "Unconfirmed". The log files that were
requested have very similar content to those which were already posted. The
issue doesn't seem to be logged. The problem is definitely with Chromium
since after switching to the Chromium based Opera browser I'm seeing the
exact same problem with Session Storage generating excessive I/O.

Again, while the attitude of some people complaining here might not be
productive, this is a serious bug that needs to get fixed. The kind of
treatment this serious issue is getting isn't helping with people's
frustration.

[chro...@googlecode.com]

Comment #69 on issue 176727 by cmum...@chromium.org: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

Did some looking into this yesterday. I used Process Monitor to record a 24
hour period of Chrome I/O. I did use Chrome a bit, but it was mostly idle
and I'm guessing that > 95% of this I/O is from background "stuff" -
extensions, etc.

I have three Chrome profiles (open in three windows), and a total of 12
open tabs. Chrome wrote just under 1GB in a day.

Cache : 310148277 (29.8%)
IndexedDB : 304667310 (29.2%)
Local storage : 96157268 (9.2%)
Local State : 88937498 (8.5%)
Preferences : 81015755 (7.8%)
Safe Browsing : 49219394 (4.7%)
Cookies : 48291856 (4.6%)
Other : 14327454 (1.4%)
PNACL : 12468290 (1.2%)
Extensions : 11508354 (1.1%)
Sync Data : 7291584 (0.7%)
JumpList Icons : 6873526 (0.7%)
Sqlite temp : 3682932 (0.4%)
Temp : 3100309 (0.3%)
Index Journal : 2387304 (0.2%)
Session storage: 1182972 (0.1%)
GCM Store : 167200 (0.0%)
Media Cache : 153276 (0.0%)
Shortcuts : 124032 (0.0%)
Filesystem : 51003 (0.0%)
GPU Cache : 35840 (0.0%)
Total : 1041791434

I probably have 50 or so extensions between my profiles. I do have
some "under development" extensions installed. It is possible that some of
these do extra logging which may be causing some of this.

Also, there are enough comments on this issue regarding session storage
that I believe something is going on there. If anybody has a reproduction
for excessive session storage I/O please let me know.

Attached is the script I threw together yesterday to analyze the Process
Monitor logs and produce this output.

So I'm still digging into this problem, and will report back here when I
learn anything interesting.

Attachments:
analyze_io.py 6.8 KB

[chro...@googlecode.com]

Comment #70 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I will send you some report soon if it happen - I am avoiding use of Chrome.

Some extensions generate some i/o whatever Chrome it self can go into
madness - I removed many extensions only left one or two - now there is no
gmail and other sync possible extensions - it is not helps - still Chrome
can become crazy and make disk busy without reason.

[chro...@googlecode.com]

Comment #72 on issue 176727 by geoedges...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

Whenever I pass a certain number of tabs, it's either number of tabs or
that I'm reaching my available limits of RAM, Chrome starts to perform an
insane amount of disk READING.

I'm an insane tab abuser, I have a ton of them open at all times.

My system has an SSD and a system activity monitor visible at all times on
my status bar, and whenever this happens I see sustained disk reading in
the hundreds of megabytes per second (probably the limit of what my SSD can
do) and it keeps on going, with my system being TOTALLY unresponsive until
Chrome decides to kill a few tabs.

It's important to note that my system is Linux, and has swap totally
disabled.

I can reproduce it very easily when needed by simply opening a lot more
tabs.

Hopefully this helps and do let me know if I can be of assistance in
reproducing/logging this.

[chro...@googlecode.com]

Comment #73 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I am author of bug - not Google employee - I think that the most of SSD
users (probably Google employees) do not know that Chrome generates such
high I/O since it is hidden - whatever it is more dangerous - if I have HDD
I see that I have to close Chrome - since system is almost dead / all is
slow.

geoedges...@gmail.com - try to catch disk usage do proof that Chrome is
doing such high i/o - I try you suggestion with multi tab open to use
Process Monitor on Windows. I looks that if you have SSD you need to switch
off swap but also Chrome - it is something wrong since Chrome has many nice
features poisoned by some dangerous code which nobody want remove and need
to be removed.

[chro...@googlecode.com]

Comment #74 on issue 176727 by Genesis....@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

check it, this is super upsetting. I no longer use chrome due to the memory
usage, this is on my friends machine. So glad I switched.

Attachments:
crazy.JPG 55.6 KB

[chro...@googlecode.com]

Comment #75 on issue 176727 by Genesis....@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

2 hours later has this been an issue since the code was written?

Attachments:
2 hours later.JPG 88.2 KB

[chro...@googlecode.com]

Comment #76 on issue 176727 by Joshua.D...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

last one before bed 10 hours usage total

Attachments:
last.JPG 59.2 KB

[chro...@googlecode.com]

Comment #77 on issue 176727 by nagy.t...@gmail.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I have the same issue as the user in #72. Linux, Kubuntu 14.04 64 bit, SSD,
but my system don't run out of free memory. When I notice the first lags of
the system I have seconds to kill chrome, or only REISUB helps
(http://en.wikipedia.org/wiki/Magic_SysRq_key#Uses) so I can't even find
out what causes this. Chrome just unexpectedly genereates so high IO
activity that my system becomes totally unresponsive in seconds.

[chro...@googlecode.com]

Comment #78 on issue 176727 by Joshua.D...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

Indeed, machine racked up over 100 million I/O writes today. For shame.
Tested a few browsers use the chromium source and it is simply how the code
is written. Opera unfortunately switched over so that burns a very
beautiful bridge that once stood tall in bustling crowd. IE (scoffs)..
tested variations of gecko engine and safari. Safari uses more memory then
that of Gecko based browsers. Though IO writes are about the same regarding
main stream browsers. 9-12,000 writes for 1080P 4 minute lengh video where
as chrome is 300-1000 per second. The best variation of the Gecko engine
I've found and have been using on my personal machine is K-Meleon. She is
weird though a few kinks between installations/exstentions, you have play
with her to make a stable version based on your needs. Not for those who
want to just start up and go, use firefox or a variation for that like
waterfox. Am using 400MB memory have this tab open, gmail, 4 videos, a
blog, facebook and a few wikipedia pages. at 40k writes in last few hours,
a full days use is usually under 200k.

[chro...@googlecode.com]

Updates:
Labels: Hotlist-Slow

Comment #79 on issue 176727 by micha...@chromium.org: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

adding this to the hotlist-slow, the premise being that disk contention ==
slow

[chro...@googlecode.com]

Comment #80 on issue 176727 by nagy.t...@gmail.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

@79 I have SSD with SATA2, I think it's not that slow.

Yesterday I managed to get some info on this. The system killed one chrome
process, I've taken a picture of it. It says the process went out of memory
(my system has 8GB RAM with swap turned off).
After I switched back to desktop I found out that the process was the
Gmail's tab which I keep constantly open.

Attachments:
IMG_20150204_142738.jpg 794 KB

[chro...@googlecode.com]

Comment #81 on issue 176727 by ma...@chromium.org: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

For people who can reproduce it: what's the minimum set of web pages to
have open when this reproduces?

Based on the previous investigation, I suspect this is about a specific set
of web pages doing something insane, like a huge amount of LocalStorage
operations.

GMail, Google Docs?

[chro...@googlecode.com]

Comment #82 on issue 176727 by nagy.t...@gmail.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

@81 After some hours of investigation since my last post I found out that
Hangouts was the problem in Gmail. I turned it off in settings, installed
the extension from webstore and the memory use is stable since. But
hangouts extension still uses way more memory than any other tab, so I
guess it is not too well optimised (maybe has some memory leak).
As I see now the 'Google Drive background process' also using high amount
of memory.

[chro...@googlecode.com]

Comment #84 on issue 176727 by cmum...@chromium.org: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

Cezary Wagner: What I/O mode is your hard drive using? DMA or PIO mode?

https://www.ifixit.com/Wiki/Enabling_DMA_Mode_Transfers

[chro...@googlecode.com]

Comment #85 on issue 176727 by nagy.t...@gmail.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

Yes, you are right, but sorry, if you look back, my first symptom was high
I/O and only thing I knew it was related Chrome.

I hope you can figure this issue out as well.

[chro...@googlecode.com]

Comment #86 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

My i/o mode is rather DMA since PIO is obsolete and not used now in modern
disks and BIOS.

I have problems yesterday with Chrome it goes into heavy traffic mode and
all stopped so I killed Chrome whatever I have not idea what triggers this
mode and how can you reproduce it I can only measure it again but not idea
how to trigger it.

I think that is not problem with memory since other applications not
generates such high i/o - especially if they are inactive. I am using now
more Firefox since it not dangerous for my computer (not I am using not
raid device) and it is faster and use less memory - whatever I can use
Chrome if problem will be solved not not want if not have to.

I do not like also idea of integration hangout with chrome - gtalk was good
product and successor is like Win7 replaced by Win8 - I do not like fat not
working things and spending time on meditation when all is loading long
time - computer and new software is to make life easier not more complex :)

[chro...@googlecode.com]

Comment #88 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I will do it if it happens again - I was not known this option before.

[chro...@googlecode.com]

Comment #89 on issue 176727 by pavel.va...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I confirm that Chrome makes the system laggy, even if not used (just opened
in background for ~8 tabs).

[chro...@googlecode.com]

Comment #90 on issue 176727 by pavel.va...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

Here is a screenshow of Win7 resource monitor/disk tab. Also please note
that there is Dropbox is on the computer, which is also io-heavy. Together
Chrome and Dropbox makes my ssd-based system to lag again and again. And
when I run BTSync same time, the computer becomes unusable at all. Three
heroes!!!

Attachments:
chrome-io-even-in-background.png 118 KB

[chro...@googlecode.com]

Comment #91 on issue 176727 by pavel.va...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

Here is another screenshot. I just opened chrome://tracing and started it
for few seconds. Then, without switching to any tabs, I stopped it.
Please look at "Chrome IO thread" hectic life! All without switching to any
tabs.

Attachments:
chrome-io-2.png 112 KB

[chro...@googlecode.com]

Comment #94 on issue 176727 by Kurtextrem: Heavy I/O generated by chrome

during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I've noticed (unlimited) IO usage when simple cache is enabled while
watching longer videos (movies). After disabling simple cache it works
again. Is this a seperate issue / known?

[chro...@googlecode.com]

Comment #95 on issue 176727 by kdefi...@gmail.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I would hesitate to call this a bug. Chrome is reading your disk, a lot of
your disk, even when you are idle. It seems to have some interplay with
FlashPlayer. Why so much disk reads. I doubt it is a bug. More likely
virus or data mining.
The same behavior is now happening with Opera and firefox

[chro...@googlecode.com]

Comment #96 on issue 176727 by geoedges...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I can easily reproduce this every time to the point chrome, X and all my
other apps and essentially the entire workstation becomes completely
unresponsive.

Not even the clock is updated on the screen. When my task bar disk
activity monitor shows over 500MB/s of disk reading activity everything
pretty much locks up until the kernel kills some chrome processes or I
power cycle the machine.

Can anyone recommend/explain how to run the tracing system in a way that
would keep the data the devs are looking for in order to pinpoint where
this issue might be coming from?

[chro...@googlecode.com]

Comment #97 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

Try this tool it is the best - if it is Windows :)

https://technet.microsoft.com/en-us/sysinternals/bb896645.aspx

Collect statistics than see files events, networks event or cpu usage and
many more - this is very powerful you can track that is tools which I was
used to find this i/o bug. If you see what doing high i\o you can isolate
this by filtering try start from simple log all than focus on root cause.

[chro...@googlecode.com]

Comment #98 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

Internal resource monitor can be also enough or process explorer from
Microsoft. In linux it is easier I think and you can limit i\o.

[chro...@googlecode.com]

Comment #99 on issue 176727 by eelgh...@gmail.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

With my 2GB RAM Debian sid and Google Chrome unstable, I get lots of disk
shuffling when I keep a Facebook page open or when I play a Youtube video
in the full-screen mode. Even moving the mouse pointer over the
full-screen video causes frantic disk accesses without a pause.

When I opened chrome://tracing, I saw lots of controls but no traces.
Trying to click the controls without knowing much about them resulted in
striped bars titled with the page names I had in other tabs. Then I saw a
concentric chart showing "v8", "blink", "Thread sampling" circles, but I
could not figure the picture they aimed at conveying to me. The numbers
beside the thread names were between 90 and 100%. Clicking the
table "Thread sampling" resulted in heavy unstopping disk access and a
prompt to kill the tracing page.

[chro...@googlecode.com]

Comment #100 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

Chrome can collect telemetry data but it generates small i/o whatever it is
constant i/o - you can check it opening form and typing without sending -
such monitoring was present in windows version but can be cross system.

No idea what mean chrome://tracing data but I think that i/o is not
monitored in chrome or it is monitored in limited ways since they can not
found problem since 2 years - I also not found method to monitor i/o from
Chrome.

I am using now Firefox for all activities and sometimes Chrome for
development only or in special case - disk failure is big problem so better
is not allow this.

[chro...@googlecode.com]

Comment #101 on issue 176727 by cory.tr...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

Chrome writes out about 3GB an hour with 10 tabs open. Windows 7.

[chro...@googlecode.com]

Comment #102 on issue 176727 by honestdu...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

Software developer here and I get this issue all the time, I just now
reproed it in thre nest version oif chrome by simply leaving tabs open on
my computer and letting my computer go idle, hibernate, and then
unhibernate, a few times.

Basically usage is:
Use windows computer, open lots of tabs on marrowwind console commands,
hibernate computer, walk away from computer and work for a week, come back
to the computer on the weekend and unhibernate the computer hopeing to do
some old school gaming, wait a while, notice entire computer is slow
despite the 6 logical cores all running on an i7 at 3.5ghz, open windows
task manager and notice that chrome is using 100% of my disk I/O for the
marrowwind browser tabs, kill chrome, watch as disk i/o drops like a stone.

This is what I as a software developer consider a critical defect, and I
wouldn't knowingly ship any code to customers with such an issue. I
honestly cant help but call into question the development practices of
Google or the Chrome team when I see defects like this repeatably over and
over again reported for years at a time with no regression tracking or
automated mitigation. After all, isn't that we as professionals are
supposed to do?

Simply put Chrome should NOT be writing to disk at all when its idle.
There is no reason for a browser that has finished loading all of its tabs
a WEEK ago to be using 100% disk i/o until the tabs are closed, and while I
fully understand you guys are running A/B tests and small bits of
processing in peoples browsers (based on the process data its easy to tell)
and despite the fact I hate this and think its invasive, you guys should at
least not be killing my hard drive.

[chro...@googlecode.com]

Comment #103 on issue 176727 by honestdu...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

Software developer here and I get this issue all the time, I just now

repro-ed it in the newest version of chrome by simply leaving tabs open on

my computer and letting my computer go idle, hibernate, and then
unhibernate, a few times.

Basically usage is:
Use windows computer, open lots of tabs on marrowwind console commands,
hibernate computer, walk away from computer and work for a week, come back

to the computer on the weekend and unhibernate the computer hoping to do

some old school gaming, wait a while, notice entire computer is slow
despite the 6 logical cores all running on an i7 at 3.5ghz, open windows
task manager and notice that chrome is using 100% of my disk I/O for the

marrowwind browser tabs opened a week ago, kill chrome, watch as disk i/o
drops like a stone back to system idle numbers.

This is what I as a software developer consider a critical defect, and I
wouldn't knowingly ship any code to customers with such an issue. I

honestly cant help but call into question the development practices of the

Chrome team when I see defects like this repeatably over and over again
reported for years at a time with no regression tracking or automated
mitigation. After all, isn't that we as professionals are supposed to do?

Simply put Chrome should NOT be writing to disk at all when its idle.

Period. There is no reason for a browser that has finished loading all of

its tabs a WEEK ago to be using 100% disk i/o until the tabs are closed,
and while I fully understand you guys are running A/B tests and small bits

of processing in peoples browsers (based on the process data in chrome
using process explorer its easy to tell) and despite the fact I personally
hate this and think its invasive as I dont like the idea of you guys
turning my pc into a node for your personal super computer, you guys should
at least not be killing my hard drive with disk IO on pages opened a week
ago that I literally just un-hibernated out of.

[chro...@googlecode.com]

Comment #104 on issue 176727 by mariusma...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

wtf, here to, this is evil....

[chro...@googlecode.com]

Comment #105 on issue 176727 by mariusma...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

wtf, here too, this is evil....

[chro...@googlecode.com]

Comment #106 on issue 176727 by mariusma...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I apologize, I just reinstalled the system and is the same pattern before
putting chrome, HDD starts chewing on when cpu is idle, if I try to find
the process he stops but returns in few mins...,
if I disconnect the internet he stops, probably from windows it self some
kind of report filled...

[chro...@googlecode.com]

Comment #107 on issue 176727 by BigSerp...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I having same problem on Windows7 Pro 64bit + Google Chrome 44.0.2399.0
canary (64-bit).

After executed Chrome and it load the default Google search engine. The
Google Chrome keep generating the disk write activity.

[chro...@googlecode.com]

Comment #109 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

Google suggest that something change I experience that there is some change
but not tested. Chrome in idle can also drain battery because without
reason CPU becomes high if there is not user activity.

I tried new Chrome version as it was suggested but it not helps - I have to
use sometimes Chrome fro development but because on danger I trying to kill
it as fast as possible since I can not afford for frequent hardware repairs
or replacement.

[chro...@googlecode.com]

Comment #110 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I will never use Chrome if you will not fix overuse of cpu and i/o - you
need plan behavior Chrome it can not take over my computer to do
prefetching, useless calculation, telemetry and high i/o without reason - I
am using not only Chrome and not want waste energy and hardware - it is
again ecology and my need as programmer.

I can not believe that is not possible to control program and resource
usage - maybe some option should be switch on on user risk if user sign
some agreement - i.e. I want to damage my disk since speed is more
important for me that my hardware. I will not sign such agreement and there
will be no problem for me.

[chro...@googlecode.com]

Comment #111 on issue 176727 by honestdu...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

It should be running I'm getting this issue on multiple versions of windows:

Windows 7
Windows 8
Windows 8.1 Enterprise

[chro...@googlecode.com]

Comment #112 on issue 176727 by marino.peter: Heavy I/O generated by chrome

during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

This bug has been there forever.. and apparently google won't fix it. In
the last couple of years I have noticed that google is being more and more
evil, which is going against their own moto "Don't be Evil".

WHY!

[chro...@googlecode.com]

Comment #113 on issue 176727 by laboboL...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I'm still having this issue.
Windows 7 Pro 64-bit
Chrome Version 43.0.2357.130 m

Chrome humps the hard-drive even when the browser sits idle in the
background. Every few seconds it access the hard-drive. If I leave the
computer on for a few hours when I come back it is still accessing the
hard-drive in the same pattern.

What is it it needs to update for hours, every few seconds, while in the
background while I'n not interacting with it?
Chrome killed my SSD hybrid. I recommend to my customers to use Firefox if
they use SSD. I'm currently using a normal spinning hard-drive.

Chrome do not let it sleep, spin-down, ever!

If you want me to create a Process Monitor log, just tell me. But I will
only send it privately. I like my privacy.

[chro...@googlecode.com]

Comment #114 on issue 176727 by masterc...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

This is getting bad for me also 64 gig RAM 6x SSD drives raid 0 ubuntu if I
use 4x windows with all tabs used my io goes through the roof

[chro...@googlecode.com]

Comment #115 on issue 176727 by masterc...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

hey guys I just want to mention I did this and now I can open more tabs
with out the pc becoming unresponsive from IO

seems you need to disable cache somehow ...


SHIFT+CTRL+I

Click on the settings icon on the right hand side looks like a cog
( Show Draw / Settings / Dock to window / X )

Click Disable cache ( while DevTools is open)

unfortunately you need to keep that dev window open just create a new tab
or window and use as normal

[chro...@googlecode.com]

Comment #117 on issue 176727 by eelgh...@gmail.com: Heavy I/O generated by

chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I see the disk shuffling gone for more than 2 months. I guess upgrading my
Linux kernel fixed the disk incessant shuffling when browsing Facebook,
watching Netflix and even moving the mouse pointer across the screen.

[chro...@googlecode.com]

Updates:
Blockedon: chromium:382272

Comment #121 on issue 176727 by cmum...@chromium.org: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

(No comment was entered for this change.)

[chro...@googlecode.com]

Comment #123 on issue 176727 by pdproduc...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I'm running a new Linux installation (Linux Mint Rafaela) and had the
Dropbox extension for Chrome activated because I use it on a different
machine. Disabling the Dropbox extension seems to have alleviated my
problem with this.

[chro...@googlecode.com]

Comment #124 on issue 176727 by honestdu...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

This is still happening; The chrome team clearly has no clue what software
craftsmanship looks like if this is still an issue and STILL have no way
to turn off all disk usage. There is simply NO REASON to be using 100% of
the hard drive AT ALL when I have a 32GB of RAM to use that just isn't
being touched by Chrome.

[chro...@googlecode.com]

Comment #125 on issue 176727 by Cezary.W...@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

Indeed it is still not solved I run lastest version to test internet
application and observe high i/o activity and some background i/o when I am
not using Chrome long time.

Chrome is flash disk killer fortunately i do not have flash disk (or
unfortunately) and Chrome only makes my computer slow - I can not use since
it is still not fixed. Prefer Firefox which use less resources and power -
it reduce usage of battery and disk.

Why do you not want remove writing on disk in idle state or when I just
typing letters in form I have not idea. You can debug and determine root
cause adding hooks on all i/o primitives then remove components which
should use RAM not disk.

[chro...@googlecode.com]

Comment #126 on issue 176727 by contact....@gmail.com: Heavy I/O generated

by chrome during inactivity - what happen/some craches?
https://code.google.com/p/chromium/issues/detail?id=176727

I'm facing this issue too. I've SSD and every day I see about 2-3GB data
gets read/written in couple of hours, just sitting chrome in idle. This is
forcing me to uninstall chrome.. Better, I should go with IE...

Attachments:
Untitled.png 112 KB