Re: Cimprovagt idle timeout
Devchandra L Meetei <[email protected]> Tue, 30 Jan 2018 15:52:59 +0530
| Newsgroups | gmane.network.open-pegasus.general |
|---|---|
| Message-ID | <CAKmFMKfSNw-nbeA2af4yEs=JVL-g3B-Bj=zV8PL-XR4E234Dcw@mail.gmail.com> |
--001a113d5c6a42f3b40563fbbfad
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Here is the code snippet at src/Pegasus/Server/CIMServer.cpp .
It will be max 2 times of value passed to _monitor.run(). I don't remember
exactly
how we deduce the value 2. But you can figure out by tracing. Did the work
almost 5 years back.
The value, in millisecond, is being passed to `select` call.
void CIMServer::runForever()
{
// Note: Trace code in this method will be invoked frequently.
if (!_dieNow)
{
_monitor->run(500000);
On Tue, Jan 30, 2018 at 2:23 PM, Prajakta Phadke <[email protected]> wrote=
:
> I agree. We should not control cimprovagt. Also, fixing memory leak is th=
e
> solution. But keeping cimprovagt idle timeout configurable will be good
> defensive approach whenever needed.
> Right now, it=E2=80=99s 8 to 10 minutes. I didn=E2=80=99t find any timeva=
l or time entity
> where it is set to 8 or 10 minutes to clean up idle threads from
> ThreadPool. But it happens. Where does that magic number picked up from?
>
> Thanks
> Prajakta Phadke
>
> Prajakta Phadke
> ------------------------------
> *From:* Devchandra L Meetei <[email protected]>
> *Sent:* Monday, January 29, 2018 9:49:39 PM
> *To:* Prajakta Phadke
> *Cc:* Kirk Augustin; [email protected]
>
> *Subject:* Re: Cimprovagt idle timeout
>
> Hi
> I thought I have opened a bug ticket when I was working on Pegasus to
> introduce a config option to control this. But could not find any.
>
> Yes, your observation is correct and it is not supported as of now.
>
> On Tue, Jan 30, 2018 at 4:55 AM, Prajakta Phadke <[email protected]>
> wrote:
>
>> This won=E2=80=99t keep cimprovagt running long for my provider. It will=
avoid
>> memory growth happening due to memory leak.
>>
>> I tried setting idleConnectionTimeout. It reflects value in cimserver
>> command arguments. But it doesn=E2=80=99t affect cimprovagt.
>>
>> It will be great if I can stop cimprovagt instance for my provider.
>>
>>
>>
>> Thanks,
>>
>> Prajakta.
>>
>>
>>
>> *From: *Kirk Augustin <[email protected]>
>> *Reply-To: *Kirk Augustin <[email protected]>
>> *Date: *Monday, January 29, 2018 at 2:00 PM
>> *To: *"[email protected]" <[email protected]>, Prajakta
>> Phadke <[email protected]>
>> *Subject: *Re: Cimprovagt idle timeout
>>
>>
>>
>> First tell us why you want to cause Cimprovagt to unload?
>>
>> It hardly has much overhead, and if there is any sort of activity
>> ongoing, you most certainly would be better off having in already loaded=
.
>>
>> In fact there is no reason for it ever to unload these days, with all th=
e
>> resources computers have.
>>
>>
>>
>> Changing idleShutdownTimeout through cimconfig should not have an
>> immediate effect, but only after the next time it is loaded.
>>
>>
>>
>> Kirk & Trish Augustin
>> 1003 Mountain Valley Rd
>> <https://maps.google.com/?q=3D1003+Mountain+Valley+Rd&entry=3Dgmail&sour=
ce=3Dg>.
>> , Tijeras, NM, 87059
>> Trish Cell: 503-333-8470
>> Kirk Cell: 505-504-1658
>>
>> Home: 505-407-8337
>>
>>
>>
>>
>>
>> On Monday, January 29, 2018, 1:38:09 PM MST, Prajakta Phadke <
>> [email protected]> wrote:
>>
>>
>>
>>
>>
>> Hi,
>>
>>
>>
>> I am trying to configure cimprovagt idle timeout.
>>
>> Changing idleShutdownTimeout through cimconfig didn=E2=80=99t help.
>>
>> Cimprovagt is invoked through ProviderAgent and uses ThreadPool.
>>
>> ThreadPool cleans up idle threads at interval.
>>
>> I have observed that cimprovagt idle timeout is around 10 minutes.
>>
>>
>>
>> I want to reduce this timeout to unload cimprovagt. Is there any
>> workaround/solution?
>>
>>
>>
>> Thanks,
>>
>> Prajakta Phadke.
>>
>
>
>
> --
> Warm Regards
> --Dev
> OpenPegasus Developer
>
> "I'm one of those people that think Thomas Edison and the light bulb
> changed the world more than Karl Marx ever did,=E2=80=9D Steve Jobs
>
--=20
Warm Regards
--Dev
OpenPegasus Developer
"I'm one of those people that think Thomas Edison and the light bulb
changed the world more than Karl Marx ever did,=E2=80=9D Steve Jobs
--001a113d5c6a42f3b40563fbbfad
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
<div dir=3D"ltr">Here is the code snippet at=20
<span style=3D"color:rgb(34,34,34);font-family:arial,sans-serif;font-size:s=
mall;font-style:normal;font-variant-ligatures:normal;font-variant-caps:norm=
al;font-weight:400;letter-spacing:normal;text-align:start;text-indent:0px;t=
ext-transform:none;white-space:normal;word-spacing:0px;background-color:rgb=
(255,255,255);text-decoration-style:initial;text-decoration-color:initial;f=
loat:none;display:inline">src/Pegasus/Server/CIMServer.cpp</span>
.<div>It will be max 2 times of value passed to _monitor.run(). I don't=
remember exactly=C2=A0<div>how we deduce the value 2. But you can figure o=
ut by tracing. Did the work almost 5 years back.</div><div><br></div><div>T=
he value, in millisecond, is being passed to `select` call.=C2=A0</div><div=
><br><div><br></div><div><div>void CIMServer::runForever()</div><div>{</div=
><div>=C2=A0 =C2=A0 // Note: Trace code in this method will be invoked freq=
uently.</div><div><br></div><div>=C2=A0 =C2=A0 if (!_dieNow)</div><div>=C2=
=A0 =C2=A0 {</div><div>=C2=A0 =C2=A0 =C2=A0 =C2=A0 _monitor->run(500000)=
;</div></div><div><br></div><div><br></div><div><br></div><div><br></div><d=
iv><br></div></div></div></div><div class=3D"gmail_extra"><br><div class=3D=
"gmail_quote">On Tue, Jan 30, 2018 at 2:23 PM, Prajakta Phadke <span dir=3D=
"ltr"><<a href=3D"mailto:[email protected]" target=3D"_blank">pphadke@t=
intri.com</a>></span> wrote:<br><blockquote class=3D"gmail_quote" style=
=3D"margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div>
<div id=3D"m_4800256042833035796compose-container" style=3D"direction:ltr">
<span><span></span></span>
<div>
<div>
<div style=3D"direction:ltr">I agree. We should not control cimprovagt. Als=
o, fixing memory leak is the solution. But keeping cimprovagt idle timeout =
configurable will be good defensive approach whenever needed.
</div>
<div style=3D"direction:ltr">Right now, it=E2=80=99s 8 to 10 minutes. I did=
n=E2=80=99t find any timeval or time entity where it is set to 8 or 10 minu=
tes to clean up idle threads from ThreadPool. But it happens. Where does th=
at magic number picked up from?</div>
<div><br>
</div>
<div style=3D"direction:ltr">Thanks</div>
<div style=3D"direction:ltr">Prajakta Phadke </div><span class=3D"HOEnZb"><=
font color=3D"#888888">
</font></span></div><span class=3D"HOEnZb"><font color=3D"#888888">
<div><br>
</div>
<div class=3D"m_4800256042833035796acompli_signature">
<div>Prajakta Phadke</div>
</div>
</font></span></div><span class=3D"HOEnZb"><font color=3D"#888888">
</font></span></div><span class=3D"HOEnZb"><font color=3D"#888888">
<hr style=3D"display:inline-block;width:98%">
</font></span><div id=3D"m_4800256042833035796divRplyFwdMsg" dir=3D"ltr"><f=
ont face=3D"Calibri, sans-serif" color=3D"#000000" style=3D"font-size:11pt"=
><span class=3D"HOEnZb"><font color=3D"#888888"><b>From:</b> Devchandra L M=
eetei <<a href=3D"mailto:[email protected]" target=3D"_blank">dlmeetei@=
gmail.com</a>><br>
<b>Sent:</b> Monday, January 29, 2018 9:49:39 PM<br>
<b>To:</b> Prajakta Phadke<br>
<b>Cc:</b> Kirk Augustin; <a href=3D"mailto:[email protected]" target=
=3D"_blank">[email protected]</a></font></span><div><div class=3D"h5"=
><br>
<b>Subject:</b> Re: Cimprovagt idle timeout</div></div></font>
<div>=C2=A0</div>
</div><div><div class=3D"h5">
<div>
<div dir=3D"ltr">Hi=C2=A0
<div>I thought I have opened a bug ticket when I was working on Pegasus to =
introduce a config option to control this. But could not find any.</div>
<div><br>
</div>
<div>Yes, your observation is correct and it is not supported as of now.=C2=
=A0</div>
</div>
<div class=3D"gmail_extra"><br>
<div class=3D"gmail_quote">On Tue, Jan 30, 2018 at 4:55 AM, Prajakta Phadke=
<span dir=3D"ltr">
<<a href=3D"mailto:[email protected]" target=3D"_blank">pphadke@tintri.=
com</a>></span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex">
<div bgcolor=3D"white" lang=3D"EN-US">
<div class=3D"m_4800256042833035796m_2347356069799305089WordSection1">
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:"Ca=
libri",sans-serif">This won=E2=80=99t keep cimprovagt running long for=
my provider. It will avoid memory growth happening due to memory leak.<u><=
/u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:"Ca=
libri",sans-serif">I tried setting idleConnectionTimeout. It reflects =
value in cimserver command arguments. But it doesn=E2=80=99t affect cimprov=
agt.<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:"Ca=
libri",sans-serif">It will be great if I can stop cimprovagt instance =
for my provider.<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:"Ca=
libri",sans-serif"><u></u>=C2=A0<u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:"Ca=
libri",sans-serif">Thanks,<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:"Ca=
libri",sans-serif">Prajakta.<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:"Ca=
libri",sans-serif"><u></u>=C2=A0<u></u></span></p>
<div style=3D"border:none;border-top:solid #b5c4df 1.0pt;padding:3.0pt 0in =
0in 0in">
<p class=3D"MsoNormal"><b><span style=3D"font-family:"Calibri",sa=
ns-serif;color:black">From:
</span></b><span style=3D"font-family:"Calibri",sans-serif;color:=
black">Kirk Augustin <<a href=3D"mailto:[email protected]" target=
=3D"_blank">[email protected]</a>><br>
<b>Reply-To: </b>Kirk Augustin <<a href=3D"mailto:[email protected]=
m" target=3D"_blank">[email protected]</a>><br>
<b>Date: </b>Monday, January 29, 2018 at 2:00 PM<br>
<b>To: </b>"<a href=3D"mailto:[email protected]" target=3D"_blan=
k">[email protected]</a>" <<a href=3D"mailto:pegasus-l@opengr=
oup.org" target=3D"_blank">[email protected]</a>>, Prajakta Phadke=
<<a href=3D"mailto:[email protected]" target=3D"_blank">pphadke@tintri=
.com</a>><br>
<b>Subject: </b>Re: Cimprovagt idle timeout<u></u><u></u></span></p>
</div>
<div>
<div class=3D"m_4800256042833035796h5">
<div>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
</div>
<div>
<div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-family:"Helvetica Neue"=
;,sans-serif">First tell us why you want to cause Cimprovagt to unload?<u><=
/u><u></u></span></p>
</div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-family:"Helvetica Neue"=
;,sans-serif">It hardly has much overhead, and if there is any sort of acti=
vity ongoing, you most certainly would be better off having in already load=
ed.<u></u><u></u></span></p>
</div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-family:"Helvetica Neue"=
;,sans-serif">In fact there is no reason for it ever to unload these days, =
with all the resources computers have.<u></u><u></u></span></p>
</div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-family:"Helvetica Neue"=
;,sans-serif"><u></u>=C2=A0<u></u></span></p>
</div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-family:"Helvetica Neue"=
;,sans-serif">Changing=C2=A0</span><span style=3D"font-size:11.0pt;font-fam=
ily:"Helvetica Neue",sans-serif">idleShutdownTimeout through cimc=
onfig should not have an immediate effect, but only after the
next time it is loaded.</span><span style=3D"font-family:"Helvetica N=
eue",sans-serif"><u></u><u></u></span></p>
</div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-family:"Helvetica Neue"=
;,sans-serif"><u></u>=C2=A0<u></u></span></p>
</div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-family:"Helvetica Neue"=
;,sans-serif">Kirk & Trish Augustin<br>
<a href=3D"https://maps.google.com/?q=3D1003+Mountain+Valley+Rd&entry=
=3Dgmail&source=3Dg" target=3D"_blank">1003 Mountain Valley Rd</a>. , T=
ijeras, NM, 87059<br>
Trish Cell: 503-333-8470<br>
Kirk Cell: =C2=A0505-504-1658 <u></u><u></u></span></p>
<div>
<p class=3D"MsoNormal"><span style=3D"font-family:"Helvetica Neue"=
;,sans-serif">Home: =C2=A0 =C2=A0 =C2=A0505-407-8337<u></u><u></u></span></=
p>
</div>
</div>
</div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-family:"Helvetica Neue"=
;,sans-serif"><u></u>=C2=A0<u></u></span></p>
</div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-family:"Helvetica Neue"=
;,sans-serif"><u></u>=C2=A0<u></u></span></p>
</div>
<div id=3D"m_4800256042833035796m_2347356069799305089ydpcacc1a42yahoo_quote=
d_7292414224">
<div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:"He=
lvetica Neue",sans-serif;color:#26282a">On Monday, January 29, 2018, 1=
:38:09 PM MST, Prajakta Phadke <<a href=3D"mailto:[email protected]" ta=
rget=3D"_blank">[email protected]</a>> wrote:
<u></u><u></u></span></p>
</div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:"He=
lvetica Neue",sans-serif;color:#26282a"><u></u>=C2=A0<u></u></span></p=
>
</div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:"He=
lvetica Neue",sans-serif;color:#26282a"><u></u>=C2=A0<u></u></span></p=
>
</div>
<div>
<div id=3D"m_4800256042833035796m_2347356069799305089ydpcacc1a42yiv40945804=
82">
<div>
<div>
<p class=3D"m_4800256042833035796m_2347356069799305089ydpcacc1a42yiv4094580=
482msonormal"><span style=3D"font-size:11.0pt;font-family:"Helvetica N=
eue",sans-serif;color:#26282a">Hi,</span><span style=3D"font-size:10.0=
pt;font-family:"Helvetica Neue",sans-serif;color:#26282a"><u></u>=
<u></u></span></p>
<p class=3D"m_4800256042833035796m_2347356069799305089ydpcacc1a42yiv4094580=
482msonormal"><span style=3D"font-size:11.0pt;font-family:"Helvetica N=
eue",sans-serif;color:#26282a">=C2=A0</span><span style=3D"font-size:1=
0.0pt;font-family:"Helvetica Neue",sans-serif;color:#26282a"><u><=
/u><u></u></span></p>
<p class=3D"m_4800256042833035796m_2347356069799305089ydpcacc1a42yiv4094580=
482msonormal"><span style=3D"font-size:11.0pt;font-family:"Helvetica N=
eue",sans-serif;color:#26282a">I am trying to configure cimprovagt idl=
e timeout.
</span><span style=3D"font-size:10.0pt;font-family:"Helvetica Neue&quo=
t;,sans-serif;color:#26282a"><u></u><u></u></span></p>
<p class=3D"m_4800256042833035796m_2347356069799305089ydpcacc1a42yiv4094580=
482msonormal"><span style=3D"font-size:11.0pt;font-family:"Helvetica N=
eue",sans-serif;color:#26282a">Changing idleShutdownTimeout through ci=
mconfig didn=E2=80=99t help.
</span><span style=3D"font-size:10.0pt;font-family:"Helvetica Neue&quo=
t;,sans-serif;color:#26282a"><u></u><u></u></span></p>
<p class=3D"m_4800256042833035796m_2347356069799305089ydpcacc1a42yiv4094580=
482msonormal"><span style=3D"font-size:11.0pt;font-family:"Helvetica N=
eue",sans-serif;color:#26282a">Cimprovagt is invoked through ProviderA=
gent and uses ThreadPool.</span><span style=3D"font-size:10.0pt;font-family=
:"Helvetica Neue",sans-serif;color:#26282a"><u></u><u></u></span>=
</p>
<p class=3D"m_4800256042833035796m_2347356069799305089ydpcacc1a42yiv4094580=
482msonormal"><span style=3D"font-size:11.0pt;font-family:"Helvetica N=
eue",sans-serif;color:#26282a">ThreadPool cleans up idle threads at in=
terval.
</span><span style=3D"font-size:10.0pt;font-family:"Helvetica Neue&quo=
t;,sans-serif;color:#26282a"><u></u><u></u></span></p>
<p class=3D"m_4800256042833035796m_2347356069799305089ydpcacc1a42yiv4094580=
482msonormal"><span style=3D"font-size:11.0pt;font-family:"Helvetica N=
eue",sans-serif;color:#26282a">I have observed that cimprovagt idle ti=
meout is around 10 minutes.</span><span style=3D"font-size:10.0pt;font-fami=
ly:"Helvetica Neue",sans-serif;color:#26282a"><u></u><u></u></spa=
n></p>
<p class=3D"m_4800256042833035796m_2347356069799305089ydpcacc1a42yiv4094580=
482msonormal"><span style=3D"font-size:11.0pt;font-family:"Helvetica N=
eue",sans-serif;color:#26282a">=C2=A0</span><span style=3D"font-size:1=
0.0pt;font-family:"Helvetica Neue",sans-serif;color:#26282a"><u><=
/u><u></u></span></p>
<p class=3D"m_4800256042833035796m_2347356069799305089ydpcacc1a42yiv4094580=
482msonormal"><span style=3D"font-size:11.0pt;font-family:"Helvetica N=
eue",sans-serif;color:#26282a">I want to reduce this timeout to unload=
cimprovagt. Is there any workaround/solution?</span><span style=3D"font-si=
ze:10.0pt;font-family:"Helvetica Neue",sans-serif;color:#26282a">=
<u></u><u></u></span></p>
<p class=3D"m_4800256042833035796m_2347356069799305089ydpcacc1a42yiv4094580=
482msonormal"><span style=3D"font-size:11.0pt;font-family:"Helvetica N=
eue",sans-serif;color:#26282a">=C2=A0</span><span style=3D"font-size:1=
0.0pt;font-family:"Helvetica Neue",sans-serif;color:#26282a"><u><=
/u><u></u></span></p>
<p class=3D"m_4800256042833035796m_2347356069799305089ydpcacc1a42yiv4094580=
482msonormal"><span style=3D"font-size:11.0pt;font-family:"Helvetica N=
eue",sans-serif;color:#26282a">Thanks,</span><span style=3D"font-size:=
10.0pt;font-family:"Helvetica Neue",sans-serif;color:#26282a"><u>=
</u><u></u></span></p>
<p class=3D"m_4800256042833035796m_2347356069799305089ydpcacc1a42yiv4094580=
482msonormal"><span style=3D"font-size:11.0pt;font-family:"Helvetica N=
eue",sans-serif;color:#26282a">Prajakta Phadke.</span><span style=3D"f=
ont-size:10.0pt;font-family:"Helvetica Neue",sans-serif;color:#26=
282a"><u></u><u></u></span></p>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</blockquote>
</div>
<br>
<br clear=3D"all">
<div><br>
</div>
-- <br>
<div class=3D"m_4800256042833035796gmail_signature">
<div dir=3D"ltr">
<div>Warm Regards<br>
--Dev<br>
OpenPegasus Developer<br>
<br>
"I'm one of those people that think Thomas Edison and the light bu=
lb changed the world more than Karl Marx ever did,=E2=80=9D Steve Jobs<br>
</div>
</div>
</div>
</div>
</div>
</div></div></div>
</blockquote></div><br><br clear=3D"all"><div><br></div>-- <br><div class=
=3D"gmail_signature" data-smartmail=3D"gmail_signature"><div dir=3D"ltr"><d=
iv>Warm Regards<br>--Dev<br>OpenPegasus Developer<br><br>"I'm one =
of those people that think Thomas Edison and the light bulb=20
changed the world more than Karl Marx ever did,=E2=80=9D Steve Jobs<br></di=
v></div></div>
</div>
--001a113d5c6a42f3b40563fbbfad--