Flex Lockup on 1.10.16

Steve K9ZW

@Jim NM1W - Was curious on the count, so I did a quick one.  I may have missed or double counted a person or two.

Eliminating duplicate posts by the same calls I make it 39 distinct reporters (I would venture there are more who never reported a lock up).

Of the 39 there are 9 saying some sort of lock-up happened only once, leaving 30 stations that have periodic issues. 

Of the 30 there are 5 saying they have the issue resolved (usually after help ticket) and 2 said the issue was other factors, leaving 23

Of the remaining 23 a couple sound like they may have other issues - the report is different than what others reported or sound like more typical network issues. 

But lets leave those is place and guess that for every one who reported via thread there is another who hasn't reported the problem.  Let's double up to 46

Presently there is almost 3600 community members, of which for simple analysis let's say 75% have a Flex-6000 and that they represent half of the Flex-6000 owners (maybe less because some folks have more than one).

Enough of that... you know where I am going and it proves nothing to say less than 1% of users...... even if it was only a single person the problem needs fixing.

---

What is important in the analysis is to use those reporting more than one freeze  the 30 distinct reports of Flex-6000 freezing in a similar way.

And some are finding ways to fix the problem.

Almost a quarter of those folk have found something to either fix, roll-back from Windows Updates, another problem, or something to adjust (or just luck) and now report the issue gone. 

There does seem to be some issues with SmartSDR for Windows Client having a memory leak in certain instances, and some reports are idle network loss reports which may be a different issue.

The handful that are mid-QSO are the most important of course. 

If I were in that group I would open a Help Ticket of course, that way the great folk at FRS could walk me through the fixes they found have worked for others.

If I were only have an odd one every couple weeks, I might just sit it out for the impending SmartSDR updates.

YMMV but I'd sure take FRS up on their help.

73

Steve K9ZW

Rick W7YP

I had another 6700 owner contact me off-forum to tell me that he's been stable since backing up to 1.10.15.  So, that's what I'm now running and have been for the last 17 hours.  So far it's still running and my fingers are still crossed.

Rick W7YP

Dismissing this issue as a "small incident rate" is like Ford dismissing the Explorer roll-over problem as a "small incident rate."  If you're one of them, you're mad as **** (if you're still alive) and rightly so.  ZERO DEFECTS used to be the same as MADE IN THE U.S.  Apparently that isn't a goal any longer.

Rick W7YP

A few people have also shared with me that they've had the lockups when controlling the radio with their Maestro, as well as when using their PC.  Others have posted the same to this forum.  This fact alone definitely puts the onus on FRS.  Were it only happening when controlled by PCs, it would understandably be tempting to blame users' PCs except for one thing:  Unless there's a problem in the radio's firmware or communication protocol design, a failure at the client end SHOULD NEVER RESULT IN A LOCKED UP RADIO!!!!  As a long time developer of embedded network systems, I can say that with 100% confidence.

The radio is essentially a server and what happens between it and the client (PC and Maestro) should not cause it to fail; yet, FRS has never taken responsibility and said, "You're right, it should never lock up."  Imagine the outrage an e-commerce software vendor providing shopping cart services for online vendors would face if their cart kept crashing and they told their vendor-customers, "It's a small incidence rate and is a problem in the shopper's PC."  

The attitude I'm getting from FRS bothers me as much as the problem itself.

Jim Runge

I have had the lock up with tone on two 6500s and a 6700. Using two different window 10 machines. I am also not at all impressed with Flex's response so far. To say they are busy with ham conventions and new radios is unacceptable. Fix what you have first. Maybe it is the Peter principle and they have reached a point where they are not capable of fixing it. What next?

Steve K9ZW

@Rick W7YP and @ Jim Runge You reported your first lockup recently. Has FRS been able to fix your issue from your Help Ticket? You've written you have a present resolution (a version roll back that is working for you)?

Any take that small numbers affected was my being dismissive of the problem is "cognitive dissonance " - go back and reread what I wrote. Maybe even read it out loud. The statement "...even it it was only a single person the problem needs fixing," is pretty clear.

The very small experience rate and the hold-back of information by not doing Help Tickets does make it harder on FRS to fix the problem. Simple facts of life are some folks have gotten interim resolution from FRS though the Help Ticket system and the process that follows from it. Sounds like FRS even has provided some trial special builds as part of the process to fix the issue. Seems like a pretty good deal all around. 73 Steve K9ZW

Paul Bradbeer

Hmm..just had my first lock-up with constant tone and needing a long-press to close down.  My 6500 is on 1.10.16, running Windows 7, DAX and CAT both in use with N1MM+ and CW Skimmer.  This was mid QSO on a 6m  Es opening to Ukraine.  Fingers crossed it won't repeat; sounds like it is an issue, so hope a fix is in the offing.  Paul M0CVX

Rick W7YP

Hi, Steve (K9ZW);

My report is recent because I have only had a 6700 since mid-January.  I also only operated it briefly while I waited for some accessories to arrive.  I was also without the radio for awhile while they replaced the FPGA fans which failed early.  Once I got the radio back and the other accessories had arrived and I finished assembling my station (also updating it to 1.10.16) and configuring the various panadapters and began operating everything 24/7, the problem soon presented itself.

Before this, I had only one HF slice receiver on one panadapter and didn't use digital modes.  The new configuration had two panadapters, one for HF and one for 2m, with the former having one slice receiver and the latter having four.  I frequently operated digitally.  The longest I've run before the radio finally crashed has been 3.5 days - ONCE.  More typically it can't do 24 hours before crashing.  I've had days when it's crashed two or three times during the day.

My ticket with FRS on this is still open.

I'm now running 1.10.15, as one fellow 6700 owner claims that rolling back to that version has rendered his radio stable once again.  As of 45 minutes ago, my radio has reached the 24 hour mark running that version.  If it makes it a week, I'll be somewhat happy.  

I have never been offered a "trial build"; in fact, I keep getting the push back that it must be something about my PC and/or network.    I've never been able to get them to accept the fact that, even if I had problems in those areas, which I do not, my problems on the client side SHOULD NEVER TAKE DOWN THE RADIO.  When/if it does, such events are highlighting a problem in the radio somewhere, as it should be extremely fault-tolerant but clearly is not.

I suspect that those who thought they had resolved their problem with FRS' help either actually had a different problem or simply don't operate enough at one sitting and/or use modes which may help to bring out the problem.

My gut instinct, based on decades of developing high-reliability networking systems and without knowing much about the radio's architecture (NOTE to FRS:  publishing that information would be a nice gesture.), is that this could well be a timing-related problem due to the addition of new features in 1.10.X.  Those tend to be very random in nature and certain uses of the device tend to predispose it to experiencing the failure.  They can also be some of the toughest problems to track down and correct; however, essentially ignoring them or instructing users to waste their own time chasing ghosts on the client side is NOT the way to deal with them.

Rick, W7YP

Frederic HB9CQK

I think your remarks are spot-on Rick! I would like to see such fact based statements from FRS here, especially regarding the path forward for us users being confronted with this serious issue.

Thanks and 73, Frédéric, HB9CQK

Andy M5ZAP

Rick,
I also think your comments are spot on and reflect my thoughts entirely.

NM1W

6700 crash; stuck in xmit during wsjtx qso; ssdr lost comm with radio;
had been on about 4 hours. long button push. no tone.

Steve Gw0gei

I have had the lockup issue since the previous public version, on the beta version and this latest public release version. I followed advice and put in another help desk ticket. After a few emails I was advised that it was a recognised issue and there was nothing they could do about it until the issue was tracked down and a solution developed. So all we can do is wait and hope for a solution at present. In the meantime my 6300 has become virtually unsellable on the second hand market. Resorted to use an elderly ft2000 and ic746 in cw field day this last weekend as I couldn't guarantee to our contest group that the 6300 was going to be stable enough to use. I asked in a previous posting for flex to advise of a previous version that it would be best to roll back to - no advice forthcoming so far. If this issue is not resolved by the end of this summer I will be boxing up my 6300 and buying a secondhand k3 or ft5000mp to use in contests over the next winter season rather than my planned upgrade to a 6700 or the 6600. Steve gw0gei / gw9j. /. Gw2op/p team

NM1W

In spite of it being a known issue (so no point in beating on Tim or Dudley about it), I think its important we continue to point out when we have crashes, or dax corruptions. The issues need to get fixed...

Rick Hadley - W0FG

I've had half a dozen instances of the constant tone lockup, but I don't know if they were all since 1.10.16 came out.  Most of them have occured overnight, or at least when I'm away from the shack.  The last one happened a couple of days ago, after I had uninstalled SSDR, done the Build 1703 Windows 10 upgrade, and then reinstalled SSDR.  The rig runs 24/7 and it hasn't happened again...yet.

Steve Jones gw0gei

I have had dax corruption on my w7 i3 laptop and my new i5 w10 tower pc.

Rick W7YP

Steve, be thankful you bought the 6300 and didn't go all the way for the 6700 like some of us did.  At least you're grabbing only one ankle instead of both of them.  I had planned to use my 6700 at our club Field Day event.  Instead, I'll bring my reliable ANAN-100D (one-third the cost).  I'll get a lot of questions as to why I brought the ANAN instead of the 6700 and I'll answer them honestly.  What this is ultimately costing FRS is going to be FAR greater than the cost of adequately staffing the research into the problem today.

Rick, W7YP

k3Tim

:-)  

I'll be backpacking my 6500 and Maestro out for Field Day to some local mountain top. Since FD isn't a 'contest' per se IF it locks up not a big deal. If I were really worried about it, I'd drop back to the working version.

Tim - k3tim/6

Steve Jones gw0gei

In the uk the rsgb hf cw field day is a contest. I am going to do another full delete of ssdr, factory reset the 6300, and try again, to see what happens. If it still occurs i will roll back to an earlier version prior to 1.10. If that doesnt solve it i will trade the 6300 in for a k3 or new anan this September and **** up the poor trade in price as one of lifes experiences to learn from.

Andy M5ZAP

Steve,

I found 1.9.13 reliable. It was only 1.10.16 and the earlier Public Beta releases that were an issue.

In fact the public Beta's were terrible especially the first release. I had to roll back to 1.9.13 for all contests. In my case there was a lower freq of lockups in 1.10.16 than in the beta releases but the problem was obviously still there.

What did flex attempt to fix between the Beta releases and 1.10.16 and does that not point them to the potential issue which is still a residual problem in some instances.

As alluded to in previous posts Flex have helped people fix the issue can they please post the fixes they have implemented. Or at least state there is no fix so that discussion can be more informed.

The silence is deafening !?

Frederic HB9CQK

Hi Andy,

maybe I am wrong, but somehow I remember that the same lock-up issue was present in a release (no beta) way back and it was fixed in a subsequent release. It reminds me of issues people have in software development with configuration management, re-introducing errors that were already fixed by linking the wrong software module or component for a new release.

Yes, the silence from FRS is overwhelming!

73, Frédéric

Tim - W4TME

There have been several requests for FlexRadio to provide an update on this issue.  The issue being addressed here is the random and occasional freezing of the radio, sometimes with an audio tone, other times without, that requires a press and hold of the radio's power button to recover.

I want to assure you that FlexRadio takes customer issues with our products very seriously.  We acknowledge that there is a problem in SmartSDR v.1.10.16 and we are committed to fixing it.  

However, this issue is a low incidence problem when compared to all FLEX-6000s in operation which directly affects the rate at which we can issue a fix.  Because of the low incident rate, it makes the problem much harder to duplicate, isolate and resolve.  We have attempted to capture data on several units experiencing these issues without success.  Whether the logging process actually affects the issue or whether the incidence rate is just too low is inconclusive.  

We continue to troubleshoot this problem and as soon as we can drive it to resolution we will issue a v1.x release that provides a fix.  But with incomplete forensics on the problem, it is hard to predict when this might occur.  

At this time the only recommendation we can suggest is to have you revert to a release that provides a stable operating environment until a fix can be delivered.  

We hear you and we are on it.   Thank you.

James Skala

Tim, We know FRS is on it.  Thanks for the update.  Keep up the GREAT work.  "Your friend and Comrade" James Skala KY4JLS

David Livingston

James Skala

See Official response. https://community.flexradio.com/flexradio/topics/flex-lockup-on-1-10-16?utm_source=notification&utm_medium=email&utm_campaign=new_reply&utm_content=topic_link I suggest you downgrade, use ddutil for spe control and allow FRS to do their part in fixing the current version.

Paul M

Tim - Can you say if there are any Alpha or Beta testers that have been having the quitting issue with 1.16,  that are now using your prerelease of 2.0, and if they are experiencing the same problems?   I know you have a Do Not Comment policy, but in this case I wonder if you can share any issues that are carried over from version 1 to version 2 related to the sudden stop issue.

David Livingston

Sorry, but I did not buy this radio to go backwards. I have been fighting this for a very long time. I believe this problem is much more common than they admit. I have at least three (3) other friends in the Columbia area having the same issue and have never made a comment about it.  I certainly hope the release of V2.??? corrects this very serious lockup issue.

NM1W

I wish they would address the DAX corruption as well..

Rick W7YP

It would be helpful if everyone who has seen a radio lockup on 1.10.X releases would post that fact to this forum; and, if you have friends who've experienced it, ask them to post the details as well.  We need to document that this isn't a rare event, as I strongly suspect it isn't.  I know that problems often go unreported for various reasons; however, failing to report isn't helping yourself OR the larger FRS community.  Remember:  It's the squeaky wheel that gets the grease.

Rick W7YP

Just made it to 72 hours of continuous operation on 1.10.15.  3.5 days is the longest I managed to do previously on 1.10.16.  Keep your fingers crossed.

KS0CW

Havent had a lockup incident in 5weeks and counting...