Closed Bug 843117 Opened 12 years ago Closed 12 years ago

[OTA][Cost Control] No data usage displayed and non-functional buttons in notification area

Categories

(Firefox OS Graveyard :: Gaia::Cost Control, defect)

ARM
Gonk (Firefox OS)
defect
Not set
normal

Tracking

(blocking-b2g:-, b2g18+ affected)

RESOLVED WORKSFORME
blocking-b2g -
Tracking Status
b2g18 + affected

People

(Reporter: whimboo, Assigned: mai)

Details

Attachments

(2 files)

+++ This bug was initially created as a clone of Bug #828298 +++ The data usage information in the notification area is not displayed after an OTA update. The buttons are non-functional and you cannot open the cost control application. An additional restart is necessary to get it working. Steps: 1. Grab a build from the day before 2. Turn on Wifi 3. Download and install the update 4. Enter PIN 5. Slide down the notification area 6. Observe the data usage fields With step 6 you will see empty fields which do not react on a tap. Not sure if it is a b2g blocker or not. But adding request for triage.
Henrik: Can you please identify which build you updated to? Are you still running on the beta channel or on the V1 train?
Hm, update channel is set to beta. So if that's not the right one, how do I change it to work on v1 builds?
Triage: QAWANTED to understand if this is reproducible since this seems to be a common issue, move to tEF?
blocking-b2g: shira? → tef?
Keywords: qawanted
I was told that cost control has changed its IDB schema and OTA updates aren't enough to bring that change in and you'll need to flash. Henrik, if you flash your phone (not just OTA update), does control control start working?
Flags: needinfo?(hskupin)
(In reply to Andrew Overholt [:overholt] from comment #4) > I was told that cost control has changed its IDB schema and OTA updates > aren't enough to bring that change in and you'll need to flash. FWIW, I can't find a bug on this ATM.
Unable to repro issue on build 20130221070203 Kernel Date: Dec 5 Gecko: http://hg.mozilla.org/releases/mozilla-b2g18_v1_0_1/rev/effe37a77f18 Gaia: bb633c6f2deb829b2f3d5b9e7bac7fa24467d02a
I flashed to todays : Gecko http://hg.mozilla.org/releases/mozilla-b2g18_v1_0_1/rev/effe37a77f18 Gaia bb633c6f2deb829b2f3d5b9e7bac7fa24467d02a BuildID 20130221070203 Version 18.0 Unagi I'm not seeing the wi-fi data move at all after playing with both wifi and mobile data. Regardless of comment 0. Something is busted I believe.
Ah. I found out that you have to change the date in order to see the cumulative number for the day. Henrik, can you change the date to tomorrow to see if the numbers change?
It has to do with initialization I believe. 1) Flash the device 2) go through FTE without setting up wifi/mobile data 3) launch cost control 4) go through cost control setup 5) turn wifi/mobile data on 6) play on the net. It won't change the numbers until you move it to the next day.
This should not happen. Data usage is updated as soon you show the utility tray or re-show the application. Check the numbers. Are the numbers below the chart increasing? If this is the case but you don't see the char increasing check the chart Y scale. Maybe it is huge compared with the current consumption (try to lower the limit).
(In reply to Andrew Overholt [:overholt] from comment #4) > I was told that cost control has changed its IDB schema and OTA updates > aren't enough to bring that change in and you'll need to flash. If that would be the case why does an additional restart help here? If OTA updates wouldn't apply the schema update, it should still not work after any further restart. So I do not think that's the problem here. Beside that usual updates have to update db schemas, or how would end users have to upgrade to a new Firefox OS release? We cannot force them to re-flash there device and to loose all their data. (In reply to Naoki Hirata :nhirata from comment #7) > I'm not seeing the wi-fi data move at all after playing with both wifi and > mobile data. Regardless of comment 0. Something is busted I believe. Naoki, if you still see this please file as a separate bug. It's not related to this one. Thanks.
Flags: needinfo?(hskupin)
(In reply to Henrik Skupin (:whimboo) from comment #11) > Beside that usual updates have to update db schemas, or how would end users > have to upgrade to a new Firefox OS release? We cannot force them to > re-flash there device and to loose all their data. Of course, you're totally right. If we file a bug, I could add a DB version manager to ensure compatibility between versions. Although, the version update took place long time ago, maybe I'm missing returning the default value if the key is not present in DB. Let's see.
(In reply to Henrik Skupin (:whimboo) from comment #11) > (In reply to Andrew Overholt [:overholt] from comment #4) > > I was told that cost control has changed its IDB schema and OTA updates > > aren't enough to bring that change in and you'll need to flash. > > If that would be the case why does an additional restart help here? Sorry, that wasn't clear to me. > Beside that usual updates have to update db schemas, or how would end users > have to upgrade to a new Firefox OS release? We cannot force them to > re-flash there device and to loose all their data. Users won't ever be on an older-than-initial-release schema so we don't need to block on that (although that would be good to fix :). (In reply to Salvador de la Puente González [:salva] from comment #13) > Although, the version > update took place long time ago, maybe I'm missing returning the default > value if the key is not present in DB. Let's see. Salva, please re-nom if you find the root cause here.
blocking-b2g: tef? → -
(In reply to Salvador de la Puente González [:salva] from comment #13) > Of course, you're totally right. If we file a bug, I could add a DB version > manager to ensure compatibility between versions. Although, the version > update took place long time ago, maybe I'm missing returning the default > value if the key is not present in DB. Let's see. Not sure what you mean here. Why would something like that show a difference if I test it right after the update or after an additional restart of the phone?
whimboo, I can file a new bug, that's fine. Having said that if the variables aren't initialized correctly, wouldn't that be the same root cause? Or are you showing numbers in the cost control app? When I said that the numbers won't change... there are no numbers after you just flash. The values aren't initialized correctly.
Bug 846402 has been filed. Also to note, are you on production or engineering build?
I can only reproduce the screenshot from comment 12 if Cost Control/Usage is not setup, ie not initialized....
I don't initialize Cost Control or even start it up. Steps I do are: 1. Update the device 2. Check data information in notification area 3. Reboot device 4. Check data information in notification area With step 2 nothing can be seen while the correct values are present in step 4.
(In reply to Henrik Skupin (:whimboo) from comment #20) > I don't initialize Cost Control or even start it up. Steps I do are: > > 1. Update the device > 2. Check data information in notification area > 3. Reboot device > 4. Check data information in notification area > > With step 2 nothing can be seen while the correct values are present in step > 4. After rebooting, are the counters increasing properly?
Hey guys. Check dates carefully. Currently Cost Control has a bug related with dates. We have found weird behaviours in some devices that, by error, change its date to a past date after the FTE.
Yes, after a second reboot the counter is working properly.
Tracking since this would be a nice-to-have, rebooting to get it to work is not an ideal workaround for the issue.
Can we verify this? Or, at least, see if it reproduces in v1.0.1?
Flags: needinfo?(carlos.martinez)
Hi all, I´ve tried this in v1.0.1 (Gecko-fe6592b.Gaia-b4d917d) and updating through OTA. Seems like it´s working fine both without setting the date / TZ correctly and setting them. We´ll test again once https://bugzilla.mozilla.org/show_bug.cgi?id=846402 has landed to double check we´re not able to repro.
Flags: needinfo?(carlos.martinez)
Hi all, I´ve tested again as bug 844774 has already landed, with v1.0.1 build (Gecko-af40bfd.Gaia-aff876b) and update through OTA and it´s working fine. Closing as worksforme.
Status: NEW → RESOLVED
Closed: 12 years ago
Keywords: verifyme
Resolution: --- → WORKSFORME
This is not fixed. I tested this with yesterdays build. Buttons are still empty and do not react on user input.
Status: RESOLVED → REOPENED
Resolution: WORKSFORME → ---
Carlos, can you verify imitating we were in a timezone GMT *minus* something? New York, San Francisco should suffice. If it works, can you provide us with a up to date STR. Thanks a lot.
Flags: needinfo?(carlos.martinez)
Keywords: verifyme
Tested in unagi device with Gecko-a378fd4.Gaia-79f122d from V1.0.1 updating to latest version and with Timezone set to América/Los Ángeles and works fine for me.
Flags: needinfo?(carlos.martinez)
Keywords: verifyme
Please Henrik, can you provide the specific way to reproduce this issue. Try to be precise (don't worry about lots of steps) because I'm not able to reproduce.
Flags: needinfo?(hskupin)
My steps as given in comment 0 should still apply. I always see this bug after an upgrade has been applied and I entered my pin. A follow-up restart makes the issue go away until the next upgrade is getting applied.
Flags: needinfo?(hskupin)
Attached file sqlite database
As attached the SQlite database for the cost control application. There is not that much data in there and it is on version 1.
Ok, let's summarize what we have here: Henrik did not flash since the working week in Berlin so it passed the FTE of Cost Control in January. Since then, he has been updating from day to day in v1-train and every time he did an update, the Cost Control widget stops to work and after a reboot, it works again. So, I'm unable to reproduce initial conditions. We have clear data for Usage with the hope the problem solves. If it does, then it was related with the initial data since, he is updating day after day, we have the same updated v1-train code Gaia. If this is not solved, probably it is related with the Gecko as if no FOTAs were installed, he has a very old back end and we can not ensure this is a Usage problem. I was trying with most recent Geckos and v1-train, and this is not reproducing.
I hit the same problem again after flashing my device today. I have not exchanged the SIM cards yet, even not run an upgrade. I just ran the FTU twice. That might be the cause why it's broken? I will check which steps are really necessary to get into this state.
Status: REOPENED → NEW
So the problem here is the dogfood-setup.sh script, which we usually have to run after flashing the device as given here: https://etherpad.mozilla.org/b2g-testdrivers-flashing Just follow the steps for flashing and updating the phone with the asset tag. Once the phone has been rebooted the cost control notification area is busted.
(In reply to Henrik Skupin (:whimboo) from comment #36) > So the problem here is the dogfood-setup.sh script, which we usually have to > run after flashing the device as given here: > > https://etherpad.mozilla.org/b2g-testdrivers-flashing > > Just follow the steps for flashing and updating the phone with the asset > tag. Once the phone has been rebooted the cost control notification area is > busted. Can we verify this and evaluate the possibility for a normal user to reproduce?
Flags: needinfo?(carlos.martinez)
Keywords: qawanted
(In reply to Salvador de la Puente González [:salva] from comment #37) > (In reply to Henrik Skupin (:whimboo) from comment #36) > > So the problem here is the dogfood-setup.sh script, which we usually have to > > run after flashing the device as given here: > > > > https://etherpad.mozilla.org/b2g-testdrivers-flashing > > > > Just follow the steps for flashing and updating the phone with the asset > > tag. Once the phone has been rebooted the cost control notification area is > > busted. > > Can we verify this and evaluate the possibility for a normal user to > reproduce? Could NOT reproduce the issue. Manually flash updated to: Inari_2013-05-23-07-02-06 nightly build. Followed steps as described in Description. OTA to: Inari Build ID: 20130524070209 Mozilla RIL Channel: Nightly Gecko: http://hg.mozilla.org/releases/mozilla-b2g18/rev/c4378ff5d057 Gaia: cc2fd02fd461aa12c96e02229a78293365d65264 After pull down of Notification, was able to see the "Used" and "limit" with actual numbers there. Upon tapping that horizontal bar, was brought to the Usage graph.
Keywords: qawanted
So, here we have tested the situation repeatedly and we are not able to reproduce. In order to avoid another re-open, asking to :whimboo, can we close this issue as WORKSFORME?
Flags: needinfo?(carlos.martinez) → needinfo?(hskupin)
Looks like that something fixed that in the past couple of days. I cannot reproduce it myself anymore. So yes, lets close it now. Thanks.
Status: NEW → RESOLVED
Closed: 12 years ago12 years ago
Flags: needinfo?(hskupin)
Resolution: --- → WORKSFORME
So the problem shows up again today. I was on vacation for a week and had my phone switched off. Not sure if that is related. But now the notification area shows the exact same thing again. After a restart I usually see bug 890824 now, but at some point this bug shows up.
Status: RESOLVED → REOPENED
Resolution: WORKSFORME → ---
Please, check the reproducibility conditions and provide more information. What is the new version of Gecko / Gaia?
Ok, I did some digging into and found the reproduction steps (at least for my case). It might still be that bug 890824 is also involved here. Steps: 1. Follow the steps on bug 890824, so that you see the locked SIM message in the notification area 2. Make a phone call 3. Pull down the notification area again After step 3 the broken UI will be present. This can be seen with: Gecko http://hg.mozilla.org/releases/mozilla-b2g18/rev/8c0127cc7be3 Gaia 7c40bdaeaffae708342fc773926dcfac5389348e BuildID 20130708070214 Version 18.0
Hm, it might be that what I see right now is related to bug 889635, which hasn't been landed on b2g18 yet. I will re-check this bug once that happened.
Please, :whimboo, bug 889635 is marked as b2g18 verified, can you double check this?
Flags: needinfo?(hskupin)
Sorry, but I don't use a b2g18 build anymore. My phone is running 1.2 now. But I haven't seen this problem for a really long time.
Flags: needinfo?(hskupin)
I´ve tested and it´s working fine. Feel free to reopen this bug if you can reproduce it. Closing as worksforme.
Assignee: nobody → mri
Status: REOPENED → RESOLVED
Closed: 12 years ago12 years ago
Resolution: --- → WORKSFORME
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: