<div dir="ltr"><div>Bad</div><div><br></div><div>root@storage1:/root# dtrace -s tcp_tput.d</div><div>^C</div><div><br></div><div> unacked(bytes) 10.250.0.3 2049</div><div> value ------------- Distribution ------------- count</div><div> 32 | 0</div><div> 64 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ 11</div><div> 128 | 0</div><div><br></div><div> unacked(bytes) 10.250.0.2 2049</div><div> value ------------- Distribution ------------- count</div><div> -1 | 0</div><div> 0 |@@@@@@@@ 63</div><div> 1 | 0</div><div> 2 | 0</div><div> 4 | 0</div><div> 8 | 0</div><div> 16 | 0</div><div> 32 | 0</div><div> 64 |@ 5</div><div> 128 |@@@@@@@@@@@@@@@@@@@@@@@ 195</div><div> 256 |@@@@@@ 54</div><div> 512 |@@ 13</div><div> 1024 |@ 5</div><div> 2048 | 0</div><div><br></div><div> unacked(bytes) 10.255.0.55 445</div><div> value ------------- Distribution ------------- count</div><div> -1 | 0</div><div> 0 | 7</div><div> 1 | 0</div><div> 2 | 0</div><div> 4 | 0</div><div> 8 | 0</div><div> 16 | 0</div><div> 32 | 10</div><div> 64 | 48</div><div> 128 | 9</div><div> 256 | 0</div><div> 512 | 0</div><div> 1024 | 1</div><div> 2048 | 0</div><div> 4096 | 0</div><div> 8192 | 0</div><div> 16384 | 2</div><div> 32768 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ 46626</div><div> 65536 | 6</div><div> 131072 | 0</div><div><br></div><div> SWND(bytes) 10.255.0.55 22</div><div> value ------------- Distribution ------------- count</div><div> 16384 | 0</div><div> 32768 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ 1</div><div> 65536 | 0</div><div><br></div><div> SWND(bytes) 10.250.0.3 2049</div><div> value ------------- Distribution ------------- count</div><div> 32768 | 0</div><div> 65536 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ 17</div><div> 131072 | 0</div><div><br></div><div> SWND(bytes) 10.250.0.2 2049</div><div> value ------------- Distribution ------------- count</div><div> 131072 | 0</div><div> 262144 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ 369</div><div> 524288 | 0</div><div><br></div><div> SWND(bytes) 10.255.0.55 445</div><div> value ------------- Distribution ------------- count</div><div> 16384 | 0</div><div> 32768 | 367</div><div> 65536 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ 49052</div><div> 131072 | 0</div><div><br></div><div>Good</div><div><br></div><div>root@storage1:/root# svcadm restart smb/server</div><div>root@storage1:/root# dtrace -s tcp_tput.d</div><div>^C</div><div><br></div><div> unacked(bytes) 10.255.0.55 22</div><div> value ------------- Distribution ------------- count</div><div> 32 | 0</div><div> 64 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ 1</div><div> 128 | 0</div><div><br></div><div> unacked(bytes) 10.250.0.3 2049</div><div> value ------------- Distribution ------------- count</div><div> 32 | 0</div><div> 64 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ 11</div><div> 128 | 0</div><div><br></div><div> unacked(bytes) 10.250.0.2 2049</div><div> value ------------- Distribution ------------- count</div><div> -1 | 0</div><div> 0 |@@@@@@@ 23</div><div> 1 | 0</div><div> 2 | 0</div><div> 4 | 0</div><div> 8 | 0</div><div> 16 | 0</div><div> 32 | 0</div><div> 64 |@ 4</div><div> 128 |@@@@@@@@@@@@@@@@@@@@@@@@@ 84</div><div> 256 |@@@@ 12</div><div> 512 |@ 4</div><div> 1024 |@@ 6</div><div> 2048 | 0</div><div><br></div><div> unacked(bytes) 10.255.0.55 445</div><div> value ------------- Distribution ------------- count</div><div> -1 | 0</div><div> 0 | 9</div><div> 1 | 0</div><div> 2 | 0</div><div> 4 | 0</div><div> 8 | 0</div><div> 16 | 0</div><div> 32 | 10</div><div> 64 | 50</div><div> 128 | 9</div><div> 256 | 0</div><div> 512 | 0</div><div> 1024 | 1</div><div> 2048 | 0</div><div> 4096 | 0</div><div> 8192 | 0</div><div> 16384 | 0</div><div> 32768 | 1</div><div> 65536 | 2</div><div> 131072 | 260</div><div> 262144 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ 27135</div><div> 524288 | 0</div><div><br></div><div> SWND(bytes) 10.255.0.55 22</div><div> value ------------- Distribution ------------- count</div><div> 16384 | 0</div><div> 32768 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ 1</div><div> 65536 | 0</div><div><br></div><div> SWND(bytes) 10.250.0.3 2049</div><div> value ------------- Distribution ------------- count</div><div> 32768 | 0</div><div> 65536 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ 17</div><div> 131072 | 0</div><div><br></div><div> SWND(bytes) 10.250.0.2 2049</div><div> value ------------- Distribution ------------- count</div><div> 131072 | 0</div><div> 262144 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ 155</div><div> 524288 | 0</div><div><br></div><div> SWND(bytes) 10.255.0.55 445</div><div> value ------------- Distribution ------------- count</div><div> 16384 | 0</div><div> 32768 | 54</div><div> 65536 | 63</div><div> 131072 | 306</div><div> 262144 | 330</div><div> 524288 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ 210636</div><div> 1048576 |@@@@@ 28037</div><div> 2097152 | 0</div><div><br></div></div><div class="gmail_extra"><br><div class="gmail_quote">On Tue, Mar 8, 2016 at 8:20 PM, Mini Trader <span dir="ltr"><<a href="mailto:miniflowtrader@gmail.com" target="_blank">miniflowtrader@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr">Running the following dtrace.<div><br></div><div><div>#!/usr/sbin/dtrace -s</div><div><br></div><div>#pragma D option quiet</div><div><br></div><div>tcp:::send</div><div>/ (args[4]->tcp_flags & (TH_SYN|TH_RST|TH_FIN)) == 0 /</div><div>{</div><div> @unacked["unacked(bytes)", args[2]->ip_daddr, args[4]->tcp_sport] =</div><div> quantize(args[3]->tcps_snxt - args[3]->tcps_suna);</div><div>}</div><div><br></div><div>tcp:::receive</div><div>/ (args[4]->tcp_flags & (TH_SYN|TH_RST|TH_FIN)) == 0 /</div><div>{</div><div> @swnd["SWND(bytes)", args[2]->ip_saddr, args[4]->tcp_dport] =</div><div> quantize((args[4]->tcp_window)*(1 << args[3]->tcps_snd_ws));</div><div><br></div><div>}</div></div><div><br></div><div><br></div><div>Is showing that the windows sizes are not going above 64k when things are not working properly.</div></div><div class="HOEnZb"><div class="h5"><div class="gmail_extra"><br><div class="gmail_quote">On Tue, Mar 8, 2016 at 7:56 PM, Mini Trader <span dir="ltr"><<a href="mailto:miniflowtrader@gmail.com" target="_blank">miniflowtrader@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr">If it helps. This doesn't happen on NFS from the exact same client. How do I file a bug?</div><div><div><div class="gmail_extra"><br><div class="gmail_quote">On Tue, Mar 8, 2016 at 1:51 PM, Mini Trader <span dir="ltr"><<a href="mailto:miniflowtrader@gmail.com" target="_blank">miniflowtrader@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr">Simple example.<div><br></div><div>1 Server 1 client.</div><div><br></div><div>Restart service everything is fast. A few hours later from same client (nothing happening concurrently) speed is slow. Restart service again, speed is fast.</div><div><br></div><div>Its like CIFS starts off fast than somehow for whatever reason if it is not used, the connection for my CIFS drives to the server becomes slow. Also this only happens when the client is downloading. Not when uploading to the server that is always fast.</div></div><div><div><div class="gmail_extra"><br><div class="gmail_quote">On Tue, Mar 8, 2016 at 1:42 AM, Jim Klimov <span dir="ltr"><<a href="mailto:jimklimov@cos.ru" target="_blank">jimklimov@cos.ru</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">8 марта 2016 г. 6:42:13 CET, Mini Trader <<a href="mailto:miniflowtrader@gmail.com" target="_blank">miniflowtrader@gmail.com</a>> пишет:<br>
<div><div>>Is it possible that CIFS will ignore TCP buffer settings after a while?<br>
><br>
>I've confirmed my systems max transfer rate using iperf and have tuned<br>
>my<br>
>buffers accordingly. For whatever reason CIFS seems to forget these<br>
>settings after a while as speed drops significantly. Issuing a restart<br>
>of<br>
>the service immediately appears to restore the setting as transfer<br>
>speed<br>
>becomes normal again.<br>
><br>
>Any ideas why this would happen?<br>
><br>
><br>
</div></div>>------------------------------------------------------------------------<br>
><br>
>_______________________________________________<br>
>OmniOS-discuss mailing list<br>
><a href="mailto:OmniOS-discuss@lists.omniti.com" target="_blank">OmniOS-discuss@lists.omniti.com</a><br>
><a href="http://lists.omniti.com/mailman/listinfo/omnios-discuss" rel="noreferrer" target="_blank">http://lists.omniti.com/mailman/listinfo/omnios-discuss</a><br>
<br>
As a random guess from experience with other network stuff - does the speed-drop happen on a running connection or new ones too? Do you have concurrent transfers at this time?<br>
<br>
Some other subsystems (no idea if this one too) use best speeds for new or recently awakened dormant connections, so short-lived bursts are fast - at expence of long-running active bulk transfers (deemed to be bulk because they run for a long time).<br>
<br>
HTH, Jim<br>
--<br>
Typos courtesy of K-9 Mail on my Samsung Android<br>
</blockquote></div><br></div>
</div></div></blockquote></div><br></div>
</div></div></blockquote></div><br></div>
</div></div></blockquote></div><br></div>