A HG612 modem is currently monitoring a connection to an ECI DSLAM.
We are noticing massive DS RSCorr errors following a few hours of connection time.
A modem reboot/resync appears to clear these errors for another few hours.
If left to run, the DS RSCorr error counts gets too high for the 2^31 (i.e. 2147483648) integer maths limit of the batch file scripts.
At that stage, cumulative DS RSCorr errors are still being reported in modem_stats.log, but as the 1 minute differences are no longer calculated, the graphs make it appear that the errors have suddenly ceased.
Can anyone using a HG612 on an ECI DSLAM confirm whether or not this also appears to be the case with your connections.
It may be an incompatibilty matter between the HG612 & the ECI DSLAM, a "fault" requiring an engineer repair or simply that ECI DSLAMS are far more rigorous in "correcting" errors rather than leaving them uncorrected & needing to be resent.
I have attached 8 days of stats from this particular connection.
The effect of Interleaving etc. being applied can be clearly seen, especially a massive reduction in errored seconds.
The far R/H side of the DS RSCorr errors graph shows where the batch file integer limit was reached & the every minute calculation ceased.
At the time of obtaining the stats, DS RSCorr errors were still mounting at the rate of around 600,000 per minute which does seem rather excessive (if it is actually being correctly reported by the HG612).
FWIW, this does NOT appear to be an issue when a HG612 is connected to a Huawei DSLAM.