meteohub.de

ever since updating to the 13-Sep build, my MB Pros show errors of the type below (it's on both of my MB Pros, connected to different consoles)
not limited to rain data - claiming suddenly that there is a read-only database or that the DB is locked.
Repeated reboots, power-cycling won't change anything

These messages are new - at least in their frequency of occurrence.
They occur in a period during which a database backup is taking place, but not only then ... also after backups are finished.
That was also so before the latest update (i.e. the backups took place and still take place at the same time) - without all these error messages ....

I have no idea why my DB should suddenly be read-only respectively locked ....
Has the DB become too large ? (765 MB)

In one of my MB Pro I moved the DB to an external (bigger) USB device - the messages remain the same.

And I please don't want to hear that I should use the internal USB because ....
It has become too small ...
Only one station but minute based data - anyhow, earlier or later the DB will reach the threshold ...
It's not a RPi where the size of the microSD card defines the threshold.

Because of this I didn't update my MB on RPi yet as I don't want to end up in a complete mess ...
If it were only a DB size issue, the imessages shouldn't be there with an external USB for the DB (128 GB).

Any advice ?


database (17.09.2021 00:02:42): cannot prepare select on read tables (after 100 tries in 9546ms): database disk image is malformed
database (17.09.2021 00:02:42): process received signal (10) in functions sigcrash < **getdiffms < read_tables_db < sync_db < *send_db_msg < *send_db_msg < *file_copy <
logger (17.09.2021 00:02:46): sensors are now providing data.
logger (17.09.2021 00:02:46): services started.
logger (17.09.2021 00:03:13): DB cannot prepare select on 'rain0total_day' (after 100 tries in 9125ms): attempt to write a readonly database (8)
logger (17.09.2021 00:03:22): DB cannot prepare select on 'rain0total_month' (after 100 tries in 9123ms): attempt to write a readonly database (8)
logger (17.09.2021 00:03:31): DB cannot prepare select on 'rain0total_year' (after 100 tries in 9124ms): attempt to write a readonly database (8)

logger (17.09.2021 00:04:14): DB cannot retrieve 'th0hum_year' (100 tries, 9089ms): database is locked (5)
logger (17.09.2021 00:04:32): DB cannot retrieve 'th0hum_day' (100 tries, 9078ms): database is locked (5)

logger (17.09.2021 00:14:22): databases connected. Update interval 3600 seconds.
logger (17.09.2021 00:14:22): data logger (version 5.2, build 14568) started.
logger (17.09.2021 00:14:23): connect station 0 (GW1000 via TCP/IP).
logger (17.09.2021 00:14:23): connect station 9 (System Data via Plug-in).
logger (17.09.2021 00:14:26): waiting for sensors (wind0, rain0, th0, thb0) to provide data.
logger (17.09.2021 00:14:26): services paused.
logger (17.09.2021 00:14:33): sensors are now providing data.
logger (17.09.2021 00:14:33): services started.
logger (17.09.2021 00:14:59): DB cannot prepare select on 'rain0total_day' (after 100 tries in 9177ms): attempt to write a readonly database (8)
logger (17.09.2021 00:15:08): DB cannot prepare select on 'rain0total_month' (after 100 tries in 9212ms): attempt to write a readonly database (8)
logger (17.09.2021 00:15:17): DB cannot prepare select on 'rain0total_year' (after 100 tries in 9226ms): attempt to write a readonly database (8)
logger (17.09.2021 00:15:26): DB cannot prepare select on 'rain0total_day' (after 100 tries in 9214ms): attempt to write a readonly database (8)
logger (17.09.2021 00:15:36): DB cannot prepare select on 'th0temp_day' (after 100 tries in 9263ms): attempt to write a readonly database (8)
logger (17.09.2021 00:15:45): DB cannot prepare select on 'th0temp_day' (after 100 tries in 9180ms): attempt to write a readonly database (8)
logger (17.09.2021 00:15:54): DB cannot prepare select on 'th0temp_day' (after 100 tries in 9264ms): attempt to write a readonly database (8)
logger (17.09.2021 00:16:03): DB cannot prepare select on 'th0temp_day' (after 100 tries in 9211ms): attempt to write a readonly database (8)
logger (17.09.2021 00:16:13): DB cannot prepare select on 'wind0avgwind_day' (after 100 tries in 9233ms): attempt to write a readonly database (8)
logger (17.09.2021 00:16:22): DB cannot prepare select on 'wind0avgwind_day' (after 100 tries in 9223ms): attempt to write a readonly database (8)
logger (17.09.2021 00:16:31): DB cannot prepare select on 'wind0wind_day' (after 100 tries in 9184ms): attempt to write a readonly database (8)
logger (17.09.2021 00:16:40): DB cannot prepare select on 'wind0wind_day' (after 100 tries in 9222ms): attempt to write a readonly database (8)
logger (17.09.2021 00:16:49): DB cannot prepare select on 'thb0seapress_day' (after 100 tries in 9252ms): attempt to write a readonly database (8)
logger (17.09.2021 00:16:59): DB cannot prepare select on 'thb0seapress_day' (after 100 tries in 9226ms): attempt to write a readonly database (8)
logger (17.09.2021 00:17:08): DB cannot prepare select on 'thb0seapress_day' (after 100 tries in 9253ms): attempt to write a readonly database (8)
logger (17.09.2021 00:17:17): DB cannot prepare select on 'thb0seapress_day' (after 100 tries in 9230ms): attempt to write a readonly database (8)
logger (17.09.2021 00:17:26): DB cannot prepare select on 'th0dew_month' (after 100 tries in 9210ms): attempt to write a readonly database (8)
logger (17.09.2021 00:17:36): DB cannot prepare select on 'th0dew_month' (after 100 tries in 9234ms): attempt to write a readonly database (8)
logger (17.09.2021 00:17:45): DB cannot prepare select on 'th0dew_month' (after 100 tries in 9255ms): attempt to write a readonly database (8)
logger (17.09.2021 00:17:54): DB cannot prepare select on 'th0dew_month' (after 100 tries in 9200ms): attempt to write a readonly database (8)
logger (17.09.2021 00:18:03): DB cannot prepare select on 'th0dew_day' (after 100 tries in 9187ms): attempt to write a readonly database (8)
logger (17.09.2021 00:18:13): DB cannot prepare select on 'th0dew_day' (after 100 tries in 9184ms): attempt to write a readonly database (8)

The problem you experience can have 2 reasons.

1. The USB stick you use did not stand the frequency of writes. You might look into dmesg (after login via ssh) to see if kernel log reports about USB stick as a total going into read-only mode which tells that it has given up.

2. By some event the database gets corrupted. This can be easily repaired if the internal stick was used, never tested on an external one. You might give it a try on "system - admin" tab to force a reboot with database rebuild. Will take time.

Even if you don't want to hear it, but I strongly recommend to use the internal stick as database. Otherwise database rebuilds and backups will not work, where an additional external stick is used. You can replace the internal 1GB by larger sizes of the same type.

ATP NANODURA SLC 4GB:
https://de.rs-online.com/web/p/usb-stic ... lsrc=aw.ds).

Thanks for the hint with the bigger internal USB - I assume it's plug and play (after opening the box).
On the other hand for that price I can get a new RPi, a 16 GB SLC microSD card and 4 years of MB license ....
And after 4 years the DB probably reaches again the 4 GB threshold.
That's for the MB Pro with the DB on the external USB.

One observation however remains - when I display the minute based charts of the past two days, there is no data missing - meaning the data is obviously saved - so the DB cannot really be read-only.
Maybe I should just ignore the messages .... (even though it makes my IT heart bleeding)

HOWEVER - I have the same messages on my 2nd MB Pro with the DB on the internal USB. And have already rebuilt the DB 2 times.
Takes long as the rebuild goes via the external USB, not because of the USB stick which works perfectly for taking backups, but because of the low USB2 port transfer speed, (which, the rebuild, would not work if it were read-only, would it ?).

But I meanwhile think that at the end of the day it's probably a dirt(-side) effect of an internal USB size issue, for both MB Pros.

I guess I will go for an alternative solution - a 8 GB microSD (SLC !) from WD is 18 EUR a piece.
I will put it in a USB adapter and connect it to one of the USB ports of the MB Pro. That's 1/7 th of the price for the 4 GB internal USB replacement (134 EUR) and will meet the SLC criterion.

Gyvate wrote: ↑Sun Sep 19, 2021 8:05 pm I guess I will go for an alternative solution - a 8 GB microSD (SLC !) from WD is 18 EUR a piece.
I will put it in a USB adapter and connect it to one of the USB ports of the MB Pro. That's 1/7 th of the price for the 4 GB internal USB replacement (134 EUR) and will meet the SLC criterion.

That should work as well.

Hey Gyvate -

Did a new usb drive solve this problem? I’m getting similar database locked errors, but my database is only ~80mb.

Ive tried a restore from backup, and force rebuild, but neither of those helped.

I also tried a larger internal SLC usb drive, and no change.

I was also wondering if maybe my database backup was corrupted, so I tried a reformatted USB, and let it make a new database, and these errors still return after maybe 12 hrs or so.

Both usb drives are ATP SLC Nanodura drives (original and new one).

So I was wondering if you were able resolve your issue?

Mine got solved - which was a mere space issue as my database had grown too big - and the 1 GB internal USB hold not only the database but a copy of the RAM database and the whole file system you see under \\METEOBRIDGE\data

I now have a real thumb drive (a USB-microSD adapter the size of a thumbnail) in which I have inserted a 8 GB SLC microSD.
It now replaces the internal USB drive and works well.

Thanks for the quick response. I was hoping the 8gb drive would resolve the problems. But I don't see any change in behavior with the larger / new drive.

Do you have the persistent 1 minute saving option enabled? I didn't have (or notice) any problems until I started making / exporting charts, and also checked this option.

yes, also 1 minute data.
I don't know what type of export you are doing and how regular this is.
Is it possible that it coincides with a backup as during backup I guess the database is locked ?
Or vice versa the system tries to execute a scheduled backup while another process has the database locked ...

I don't run the automated backup script currently. So no scheduled backups (unless the MBPro does something behind the scenes).

I was wondering if selecting the 1 min data interval box, maybe caused more frequent database writes, which like you mention, now coincide with another process? I reduced my events to an HTTP unload to the Saratoga templates every 5 sec, an SFTP upload to these same templates every 5 mins. And then a per minute-last 24 hr-chart every 5 minutes, and a per hr-last-seven-days-chart every hr.

Using these same settings / intervals, if I start with a newly formatted USB drive and blank database, i start to see the errors after ~ 12 hrs or so. If I restore the database from a backup, I usually see the errors within an hour or so.

@Wheetie: Are you running your MB PRO on an external non-SLC USB stick?

admin wrote: ↑Thu Jan 27, 2022 2:21 pm @Wheetie: Are you running your MB PRO on an external non-SLC USB stick?

No - Internal ATP Nanodura SLC USB. I started to see these problems with the original 1GB stick, and I replaced the internal with a larger ATP SLC drive to see if that would help. It did not.

Sometimes these readonly errors are also accompanied by disk I/O errors. I noticed this again last night and the internal USB drive was remounted to dev/sda, which doesn't sound correct.

database (27.01.2022 05:05:33): cannot insert into 'thb0temp_min' (after 0 tries in 1ms): attempt to write a readonly database (8)
database (27.01.2022 05:05:33): rollback on insert forced (1) (storemin): thb0temp
database (27.01.2022 05:06:00): failed with step exist record on 'th3temp_min' (after 0 retries in 1ms): disk I/O error (10)
database (27.01.2022 05:06:00): rollback on insert forced (1) (storemin): th3temp
database (27.01.2022 05:06:04): failed with step exist record on 'th3temp_min' (after 0 retries in 1ms): disk I/O error (10)
database (27.01.2022 05:06:04): rollback on insert forced (1) (storemin): th3temp
database (27.01.2022 05:06:04): process terminated normally.
test5m (27.01.2022 13:06:16): USB storage remounted: /dev/sda
system (27.01.2022 13:06:17): start of logging process initiated.
database (27.01.2022 05:06:18): process started.
logger (27.01.2022 05:06:20): databases connected. Update interval 3600 seconds.
logger (27.01.2022 05:06:20): data logger (version 5.2, build 14655) started.

As discussed via PM we will try a change of the main board of the PRO. Looks like a HW defect.