Hi Ahmed,

I Checked the server log and matched the info with the screen prints.

Indeed the trouble of the Lightstreamer server seems to start at 12:21:59. Since then, the situation of the used memory, who had been always quite good, begins to degrade at a rate of about 300 MB every 2 seconds.
This situation goes on until to force the JVM to significant GC operations to a final reallocation up to 12 GB of total heap. In these moments the server undergoes periods of freezing (few seconds) that caused the disconnection of some clients, not all.
After about half an hour, the situation seems to be returning with a regular memory consumption more appropriate.

In the log in the relevant period of the issue the only activity of some importance is the reconnection of a client (IP address: 41.209.249.5) that, at the restart, performs several subscriptions including one repeated 60 times:

Code:
LS_session=S245abe733d79e813T2157492&
LS_op=add&
LS_table=X&
LS_id=ts1%20ts2%20ts3%20ts4%20ts5%20ts6%20ts7%20ts8%20ts9%20ts98%20ts131%20ts157%20ts132%20ts137%20ts138%20ts144%20ts146%20ts91%20ts166%20ts130%20ts86%20ts153%20ts94%20ts169%20ts83%20ts87%20ts97%20ts10%20ts11%20ts12%20ts13%20ts14%20ts15%20ts16%20ts17%20ts18%20ts19%20ts20%20ts21%20ts22%20ts23%20ts24%20ts25%20ts26%20ts27%20ts28%20ts29%20ts30%20ts31%20ts32%20ts33%20ts34%20ts122%20ts35%20ts36%20ts37%20ts38%20ts39%20ts40%20ts41%20ts80%20ts127%20ts114%20ts117%20ts155%20ts158%20ts161%20ts163%20ts168%20ts42%20ts123%20ts43%20ts44%20ts45%20ts46%20ts47%20ts124%20ts152%20ts135%20ts88%20ts156%20ts162%20ts164%20ts167%20ts170%20ts48%20ts49%20ts50%20ts51%20ts52%20ts53%20ts54%20ts55%20ts56%20ts125%20ts57%20ts58%20ts59%20ts60%20ts61%20ts62%20ts63%20ts64%20ts65%20ts66%20ts99%20ts134%20ts67%20ts113%20ts107%20ts101%20ts102%20ts115%20ts116%20ts151%20ts126%20ts136%20ts142%20ts160%20ts96%20ts68%20ts69%20ts70%20ts71%20ts72%20ts73%20ts74%20ts75%20ts76%20ts121%20ts139%20ts154%20ts77%20ts165%20ts95%20ts78%20ts79%20ts103%20ts81%20ts104%20ts82%20ts100%20ts105%20ts108%20ts106%20ts111%20ts110%20ts112%20ts109%20ts118%20ts119%20ts120%20ts133%20ts129%20ts128%20ts84%20ts85%20ts141%20ts140%20ts90%20ts143%20ts89%20ts148%20ts147%20ts149%20ts93%20ts159%20ts145%20ts150%20ts92%20&
LS_mode=COMMAND&
LS_schema=TradePrice%20TradeVolume%20TradeTime&
LS_data_adapter=PROXY_MarketBeat&
LS_Snapshot=true on "Lightstreamer HTTP Server" from 41.209.249.5:20622
Is this a normal behavior of your application?

I can not be sure that this is related at the beginning of the problem, it could simply be a coincidence, and the higher memory consumption due to an increase in the flow throughput of the incoming data from the Remoto Adapter into the server.

Could you check if there has been a considerable increase in the flow of data from the adapter to the server around that time? Have you some instruments to measure it?

Thank you,
Giuseppe