You are here

Problems with Develop 160 and 161 upgrades - Bullet Code

3 posts / 0 new
Last post
k1ky
k1ky's picture
Problems with Develop 160 and 161 upgrades - Bullet Code
I've noticed some not-so smooth upgrades from Develop 159 code to 160 and 161 on Airgrid M5 nodes using the Bullet sysupgrade.  I have tried several times with only one out of 4 upgrades gone smoothly as the previous upgrades have done. The nodes never reboot and remain on the same firmware version before upgrade. They also sometimes aren't accessible with the web interface after the procedure. They also do not seem to attempt a reboot.  I'm using Chrome, which has a nice upload progress indicator in the bottom left corner of the browser screen, which "seem" to have reached the 100% mark.  These are stable RF paths and are over-the air upgrades at remote sites. 

I'm able to use Putty to log in and reboot the nodes., which still indicate that they are running the pre-upgrade firmware. Upon further tries, I find that they do upgrade smoothly if I reboot the nodes immediately before performing the upgrade procedure.  Maybe a low memory situation?  These are units that do not have tunnels or any other services installed.

At any rate, just wanted to fly this up the flagpole and see if anyone else has experienced this before we try to dive into it deeper.
 
KG6JEI
Do you have a support data
Do you have a support data file from the node from when it fails? (Basically anytime you have an issue plan in getting a support data file is a standard required item to be investigated)

Also best to log it into bloodhound which is the official reference system for any issues. If it isn't in bloodhound it may never be touched.
k1ky
k1ky's picture
No support data yet
No support data as of yet as I know the development work is ongoing and wanted to see if this was isolated to my system for some reason before submitting a ticket.. We are going to work the whole "upgrade" issue this week regarding the upgrade of tunneled units and gather the much needed support data and will gather info on this as well.  Most of our evaluators are spread out away from their home systems for the holidays.  This seems to be a "new" issue just discovered over the past few days with the introduction of Dev 160 - 161. I'm thinking that we should gather a support data file "before" and "after" the failed load.  Just flying it up the flagpole as a query before we actually cry WOLF!

Theme by Danetsoft and Danang Probo Sayekti inspired by Maksimer