Raspberry Pi - Page 5

EdH · 2016-12-19, 23:53

OK, I might have it fixed. Time to be embarrassed... When I elected to move from one script cycling via a while loop, to two scripts calling each other, so I could make changes while the other ran, I failed to let them finish after calling and set up a recursive system with no escape until I rebooted. At least that's how it looks ATM.

I'll know more tomorrow, but for now, there is no longer an appreciable lag.

Mark Rose · 2016-12-20, 00:16

Quote:

Originally Posted by EdH

OK, I might have it fixed. Time to be embarrassed... When I elected to move from one script cycling via a while loop, to two scripts calling each other, so I could make changes while the other ran, I failed to let them finish after calling and set up a recursive system with no escape until I rebooted. At least that's how it looks ATM.

I'll know more tomorrow, but for now, there is no longer an appreciable lag.

I will admit to having done the same in the past.

EdH · 2016-12-20, 02:29

Well, unfortunately, time has shown that it is not fixed. Now, after running well for several hours, it won't talk to me for about 25 seconds, even after the scripts have been interrupted.

In the earlier top runs, I'm assuming the sshd entries are from the clients checking for assignments, but I don't know what all the lsb_release entries are for. lsb appears to reference Linux Standard Base, but are they also something to do with the client requests?

Is there something in the interrupts that stands out as overloading, with all the high numbers in the troubled watch reference?

Thanks...

Mark Rose · 2016-12-20, 03:08

Seeing all those lsb_release commands running at once seems odd to me. It would be worth while figuring out what is calling it. When you see them, can you run `pstree` when they appear, perhaps in a second terminal? You may need to install the psmisc package first.

EdH · 2016-12-20, 04:51

Quote:

Originally Posted by Mark Rose

Seeing all those lsb_release commands running at once seems odd to me. It would be worth while figuring out what is calling it. When you see them, can you run `pstree` when they appear, perhaps in a second terminal? You may need to install the psmisc package first.

pstree is available, but right now my scripts have crashed on the nfs part of a c129. I'm having to run msieve manually to finish prior to the next run of the RP3. I'll probably have to wait for the morning, since I'm not seeing many lsb_release msgs while it's idling.

xilman · 2016-12-20, 07:28

Quote:

Originally Posted by EdH

I failed to let them finish after calling and set up a recursive system with no escape until I rebooted. At least that's how it looks ATM.

Here's what the old-timers do to avoid a reboot. I'll approach it in stages so you can see what's happening.

First, find the process IDs of the things you want to stop: ps ax | grep errant_process_name | colrm 6. This will also find the grep process itself, which should be harmless but inelegant, so a grep -v grep | could be added before the colrm.

Next, take those PIDs and send them a STOP signal. This prevents them from running but doesn't terminate them so there will be no recursive calls. Note the use of back-ticks to interpolate the result of a sub-command into the command line.
kill -STOP `ps ax | grep errant_process_name | colrm 7`

Finally, terminate them with extreme prejudice, kill -KILL `ps ax | grep errant_process_name | colrm 7`, and you should now have a clean running system.

ldesnogu · 2016-12-20, 08:24

Didn't know about colrm, nice!

Regarding killing processes with their name, I use the pkill comand. Very handy.

xilman · 2016-12-20, 11:06

Quote:

Originally Posted by ldesnogu

Didn't know about colrm, nice!

Regarding killing processes with their name, I use the pkill comand. Very handy.

I'm an old-timer. colrm appeared in Unix a very long time before pkill showed up.

ldesnogu · 2016-12-20, 11:50

Quote:

Originally Posted by xilman

I'm an old-timer. colrm appeared in Unix a very long time before pkill showed up.

Indeed, I found that colrm has been written in 1974

Mark Rose · 2016-12-20, 12:01

`colrm` is also new to me. Thanks!

EdH · 2016-12-20, 15:25

Thanks guys! In my case, wouldn't just terminating the current script cause all calling ones to finally close? I hope I have that fixed anyway.

I'm no longer seeing the full screen of lsb_release processes. Only a couple show up at a time. Maybe they were linked to my script issue.

But, it's still taking more than 15 seconds to communicate with it.

I looked for bash instances via my new knowledge and there are 7 running. I can only account for 2. But, 7 shouldn't be a trouble, I wouldn't think.

I will play some more with my new tools when I have a little more time.

Thanks all.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
raspberry pi3 and srxsieve speed	pepi37	Hardware	6	2018-03-20 18:01
Runs Prime95 on Raspberry Pi	primeawesome	Hardware	6	2018-02-14 08:19
Which SIMD flag to use for Raspberry Pi	BrainStone	Mlucas	14	2017-11-19 00:59
Raspberry Pi	lavalamp	Hobbies	10	2017-08-16 00:37
Raspberry Pi	sloppyonefoot	Software	1	2017-07-02 08:48

2016-12-19, 23:53	#45
EdH "Ed Hall" Dec 2009 Adirondack Mtns 12674₈ Posts	OK, I might have it fixed. Time to be embarrassed... When I elected to move from one script cycling via a while loop, to two scripts calling each other, so I could make changes while the other ran, I failed to let them finish after calling and set up a recursive system with no escape until I rebooted. At least that's how it looks ATM. I'll know more tomorrow, but for now, there is no longer an appreciable lag.

2016-12-20, 02:29	#47
EdH "Ed Hall" Dec 2009 Adirondack Mtns 2²×13×107 Posts	Well, unfortunately, time has shown that it is not fixed. Now, after running well for several hours, it won't talk to me for about 25 seconds, even after the scripts have been interrupted. In the earlier top runs, I'm assuming the sshd entries are from the clients checking for assignments, but I don't know what all the lsb_release entries are for. lsb appears to reference Linux Standard Base, but are they also something to do with the client requests? Is there something in the interrupts that stands out as overloading, with all the high numbers in the troubled watch reference? Thanks...

2016-12-20, 03:08	#48
Mark Rose "/X\(‘-‘)/X\" Jan 2013 https://pedan.tech/ 2⁴×199 Posts	Seeing all those lsb_release commands running at once seems odd to me. It would be worth while figuring out what is calling it. When you see them, can you run `pstree` when they appear, perhaps in a second terminal? You may need to install the psmisc package first.

2016-12-20, 08:24	#51
ldesnogu Jan 2008 France 3×199 Posts	Didn't know about colrm, nice! Regarding killing processes with their name, I use the pkill comand. Very handy.

2016-12-20, 12:01	#54
Mark Rose "/X\(‘-‘)/X\" Jan 2013 https://pedan.tech/ 2⁴·199 Posts	`colrm` is also new to me. Thanks!

2016-12-20, 15:25	#55
EdH "Ed Hall" Dec 2009 Adirondack Mtns 2²×13×107 Posts	Thanks guys! In my case, wouldn't just terminating the current script cause all calling ones to finally close? I hope I have that fixed anyway. I'm no longer seeing the full screen of lsb_release processes. Only a couple show up at a time. Maybe they were linked to my script issue. But, it's still taking more than 15 seconds to communicate with it. I looked for bash instances via my new knowledge and there are 7 running. I can only account for 2. But, 7 shouldn't be a trouble, I wouldn't think. I will play some more with my new tools when I have a little more time. Thanks all.