[SOLVED] Odd Behavior of HP Printer Waking Up Every So Often: Slackware 64-bit 14.1 Stable

tronayne · 03-26-2014, 07:40 AM

Since updating packages over the past few weeks (the patches for Slackware 64-bit 14.1 stable) I've noticed that the HP Photosmart C4680 USB printer that sits next to me wakes up periodically and is doing strange things; it sounds like it's reset (the cartridge carrier activates) and, every once in a while, it switches on the scanner/copier display. It also will, sometimes, kick out a blank sheet. Never done any of this before (the network printer has always done this every so often -- like for the past 10 years -- but never the USB printer). That printer has always just sat there quietly until told to do something.

I'm wondering if anybody else has seen this and perhaps a rebuild of HPLIP from source would be worthwhile; too, there is a newer release of HPLIP than 3.13.10 that may be worth installing?

As I'm writing this it just decided (like right now) to do a color copy all by itself. Wow. Little daemons running around? Gremlins in the chips?

Tiz a puzzlement.

Chuck56 · 03-27-2014, 05:11 PM

Quote:

Originally Posted by tronayne

... I've noticed that the HP Photosmart C4680 USB printer that sits next to me wakes up periodically ...

My HP Officejet 4620 periodically "wakes up" too. It's on the LAN via wireless, not USB. Sometimes it's the printer carriage that cycles and sometimes it's the scanner that cycles. Never did find out why either.

tronayne · 03-28-2014, 08:10 AM

Yesterday, I compiled HPLIP 3.14.1 (3.14.3, the current version from H-P, won't compile with hplip.SlackBuild from the source directory), removed the printer from CUPS and reinstalled with with hp-setup. Still does the same thing, just not quite so often and there's no blank pages in the output tray... yet. Plus turning it off and removing power completely then reconnecting the power (it has one of those transformer gadgets in the power line) didn't do anything either.

The odd thing is that this is only has been happening since a batch of security patches were installed (this is Slackware 64-bit 14.1 stable). I can't imagine which of those may -- may -- have some glitch or other.

The other printer, an Ethernet-connected H-P Business Inkjet 2280tn, has done this sort of thing since day one (it cycles periodically, always has, the manual says it will). This one, though, never has (and I don't think it's supposed to, either).

So, I think, wait till it dies and get another one.

Sigh.

ljb643 · 03-28-2014, 06:24 PM

OK, here's a real long-shot. Does the printer have a memory card reader or USB socket for a flash drive? (Some HP Photosmarts do, or used to, for printing photos directly from a camera's memory card.) The reason I ask is that 'recent security patches' includes udisks and udisks2, which are know to poll storage devices regularly. (They were keeping my external USB disk from sleeping, for one thing.) You might want to run "udisksctl monitor" and let it run a while and see if any output from there corresponds to the printer doing stuff.

tronayne · 03-29-2014, 08:24 AM

Quote:

Originally Posted by ljb643

OK, here's a real long-shot. Does the printer have a memory card reader or USB socket for a flash drive?

Not such a long-shot methinks -- it does; two slots for memory cards. And, I think, this problem started showing up when those patches went on sometime after 11 March.

I'll run udisksctl monitor for a while and see what's what (if anything). Don't know diddly beans about udisksctl (yet another learning curve to climb!), but, hey, it's a place to start.

(And, mumble, grumble, maybe I'll just put the release packages back on and be done with it.)

Thanks!

MDKDIO · 03-30-2014, 02:50 PM

IF I remember it right, there's a periodical cleaning cycle on most (if not all) ink printers.
Once every now and then it'll simply cycle in order to keep the printheads clean.

Don't remember exactly how often this happens, but guessing once a week or so...

Makes sure the inc cartridges has not expired (check date on cartridges), and of course, check if using original cartridges.

Just a thought...

/MDKDIO

tronayne · 03-30-2014, 04:36 PM

Yeah, the network printer (HP Business Inkjet 2280th) does that cleaning cycle thing, always has, about once every 8 - 12 hours or so.

This printer never has -- until now -- and I'm more suspicious of udisks; again, until those patches went on this never happened. The C4680 always just went sleepy-bye and stayed that way until a job got sent to it. I'm inclined to remove the patches (I'm not real worried about the security problem) and reinstall the distribution udisks and udisks2 packages to see what happens (after I copy any config files somewhere so I can make a comparison between the old and new). As an added plus, why would the thing decide to scan on it's own (the inside of the platen cover)? It does that, by actual blank paper count, 4 times a day.

Good thinking about the cleaning cycle and cartridges, though, appreciate it.

ljb643 · 03-30-2014, 07:46 PM

I see the bug report you opened on udisks. Does this mean you have identified udisks as the cause? From the info you attached, I see the printer's card reader is detected, and "detection by polling: 1" is suspicious... The answer is probably a udev rule to disable this feature for that device, but I don't know the exact keyword (or even if there is one).

MDKDIO · 03-31-2014, 01:58 AM

tronayne
Sounds like a plan. Do let us know what you find out!
Thanks in advance!

ljb643 · 03-31-2014, 09:16 PM

Wow, that's harsh. "RESOLVED INVALID", and "don't reopen until you have something more concrete". Ouch.

Seriously, though, you should try to narrow it down. If you suspect udisks2, do that "udisksctl monitor". If it looks like that's it, add a udev rule with env{UDISKS_IGNORE}="1" for the printer's storage device, and see if the bad behavior goes away.

If that makes it go away, problem solved, but if you want to go further, you remove the udev rule, downgrade to udisks2-2.1.0, convince yourself that it is fine, upgrade back to udisks2-2.1.3, see if it comes back. Now you can do a bug report that is harder to dismiss. By the way, does the printer use up any ink when it cycles due to wakeup? If it does, and it was my printer, that would raise the bug report priority way up, given the cost of ink.

tronayne · 04-01-2014, 10:13 AM

Well, yeah, ouch (probably deserved). I know a lot of stuff but what I don't know could fill an encyclopedia, particularly about UDEV. I thought about trying to write a rule such as you suggest but then I thought about the fact that I occasionally (well, rarely) plug in the memory card from a camera to print stuff and wonder about the result of doing that: would the slot on the server still work? Would the slot on the printer still work? Do I have to fiddle and twiddle and fuss, remember the lie I told and lose function (that I may or may not care about)? Arrgghh!

This morning there are four sheets of paper in the tray from scans. Arrgghh!

The printer isn't using any ink (it's scanning the white inside of the cover) but that's not really the point, eh? It's what the hell happened and I ain't smart enough to figure that one out (yet). I'm working on it but so far no Joy in Mudville (it got into the 40's from the 10's yesterday or the day before and the four feet of snow outside is turning into water on top of frozen ground, sigh). The other printer, an Ethernet connection, doesn't have any problems (which stands to reason, it's not USB) and plotter hasn't tried to draw an E-size plot (yet); plotter is turned off just in case (that dang paper roll is expensive!).

Starting udisks --monitor and watching it is today's little challenge. See what's what. Going back to the previous versions, well, maybe, but that won't solve whatever the problem is (if it actually is udisks). Also the statement:

Quote:

Also, you should uninstall udisks 1.x - that version is no longer supported and
distributions shouldn't be shipping it.

just doesn't sound right to me -- I don't think Pat would do a double update (on both versions) without a reason.

And, of course, it could be something else entirely (but plugging the printer into a 14.1 box without all the patches does not exhibit the problem in 24 hours so I'm going to say it's probably not the printer).

Thanks for your input.

tronayne · 04-01-2014, 02:25 PM

I ran udisksctl monitor from a little before noon to at little after 15:00 (the attachment is the log of that).

udisks messes with the printer. I'm not sure how and I'm not sure what I can do about it (have some learning to do about UDEV rules).

If anybody has any ideas of what to (or a pointer to some documentation), I'd be grateful.

ljb643 · 04-01-2014, 04:33 PM

I looked at your udiskctl monitor log. Problem is, you need to correlate the times of the events with when the printed/scanner did something, and of course I can't do that.

However, I do see something bothersome. No, there is no evidence udisks is polling the device. Instead, it looks like the device is disconnecting and reconnecting. It did this at 12:42:41 and reconnected after 22 seconds. It did it again at 13:36:35 and reconnected after about 7 seconds. Again at 13:48:46. This looks to me like udisks is responding to events, not causing them.

Do these times correspond to anything in /var/log/messages or /var/log/syslog where the kernel shows a USB device disconnect/reconnect?

Do you have the printer connected through a hub? A powered hub? An extra long cable? Any other reason you can think of why the printer would reconnect?

tronayne · 04-02-2014, 08:34 AM

Quote:

Originally Posted by ljb643

I looked at your udiskctl monitor log. Problem is, you need to correlate the times of the events with when the printed/scanner did something, and of course I can't do that.

Yeah, it sits next to me -- I was working on other things with a terminal window open over in the corner of the screen running udisksctl monitor and every time entries popped up there was activity; the "TouchSmart" display popped on, the motor cycled (it did not scan during any of those and there were no blank pages this morning).

Quote:

Originally Posted by ljb643

However, I do see something bothersome. No, there is no evidence udisks is polling the device. Instead, it looks like the device is disconnecting and reconnecting. It did this at 12:42:41 and reconnected after 22 seconds. It did it again at 13:36:35 and reconnected after about 7 seconds. Again at 13:48:46. This looks to me like udisks is responding to events, not causing them.

Hm. That sorta-kinda makes a whole lot of sense -- it's monitoring what the USB ports are doing. Duh.

Quote:

Originally Posted by ljb643

Do these times correspond to anything in /var/log/messages or /var/log/syslog where the kernel shows a USB device disconnect/reconnect?

Well, yeah, the first time in the log is 12:42 and this appears in /var/log/messages:

Code:

Apr  1 12:30:33 fubar -- MARK --
Apr  1 12:42:41 fubar kernel: [103944.049029] usb 5-6: USB disconnect, device number 36
Apr  1 12:43:02 fubar kernel: [103964.679020] usb 5-6: new high-speed USB device number 37 using ehc
i-pci
Apr  1 12:43:02 fubar kernel: [103964.793867] usb 5-6: New USB device found, idVendor=03f0, idProduc
t=7411
Apr  1 12:43:02 fubar kernel: [103964.793872] usb 5-6: New USB device strings: Mfr=1, Product=2, Ser
ialNumber=3
Apr  1 12:43:02 fubar kernel: [103964.793876] usb 5-6: Product: Photosmart C4600 series
Apr  1 12:43:02 fubar kernel: [103964.793879] usb 5-6: Manufacturer: HP
Apr  1 12:43:02 fubar kernel: [103964.793882] usb 5-6: SerialNumber: CN985M155J05BQ
Apr  1 12:43:02 fubar kernel: [103964.794259] usb-storage 5-6:1.3: USB Mass Storage device detected
Apr  1 12:43:02 fubar kernel: [103964.794298] scsi46 : usb-storage 5-6:1.3
Apr  1 12:43:02 fubar logger: loading HP Device 005 037
Apr  1 12:43:03 fubar kernel: [103966.110128] scsi 46:0:0:0: Direct-Access     HP       Photosmart C
4600 1.00 PQ: 0 ANSI: 5
Apr  1 12:43:03 fubar kernel: [103966.112789] sd 46:0:0:0: [sdb] Attached SCSI removable disk
Apr  1 13:10:33 fubar -- MARK --

The same sequence appears again at 13:30 and the only thing that shows up in /var/log/syslog is the "normal" (for Slackware)

Code:

Apr  1 08:12:00 fubar hp-upgrade: hp-upgrade[27297]: warning:  distro is not found in AUTH_TYPES
Apr  2 08:11:59 fubar hp-upgrade: hp-upgrade[20571]: warning:  distro is not found in AUTH_TYPES

This is getting interesting.

Quote:

Originally Posted by ljb643

Do you have the printer connected through a hub? A powered hub? An extra long cable? Any other reason you can think of why the printer would reconnect?

Nope, connected directly to the USB port on the system with a three-foot cable. And, of course, none of the other USB devices connected; e.g., keyboard, mouse, are acting up in any way.

I'm thinking, OK, lets swap the USB cable to a different port and see (already did that actually), then find us a different USB cable and pop that in and see what happens. Or, haul out the lap top, hook it up there and see what happens. Or just disconnect the USB cable and see what happens (probably nothing without the cable attached to make the printer wake in the first place).

Or, the printer has just decided to be gone goofy in which case we go to Staples and get another one (but, dang it, the thing prints just fine).

Sigh.

Thanks for your helpful input.

ljb643 · 04-02-2014, 02:23 PM

Quote:

Originally Posted by tronayne

... Or just disconnect the USB cable and see what happens (probably nothing without the cable attached to make the printer wake in the first place).

In my opinion, the PC is not waking up the printer. I think the printer is having these little unscheduled brief naps. As much as I want to blame udisks and the timing of the security patch, I'm not seeing any admissible evidence.

Disconnecting the USB cable would be a good test, although the printer might act differently anyway with no USB cable connection.

I wonder if your printer's power supply is failing.