CBL 2.8 some devices failed to send data to sync gateway

Hello!

I am facing following issue, on some of my devices I cannot send data to server, so push replication does not work.
I am using CBL 2.8 and SG 2.8.

In SG 2.8 I see following logs:

2021-02-16T13:40:26.549+03:00 [ERR] c:[5302e508] Error during downloadOrVerifyAttachments for doc c89b2264-c56d-4d4b-afeb-9de5d0cc9f4d/4-527c4512ec85f84f87ca04868a5ee511030f44c7: 400 Incorrect data sent for attachment with digest: sha1-frvGAhWwfyN1/rO+7UMuhM5KJxM= – db.(*blipHandler).handleRev() at blip_handler.go:763
2021-02-16T13:40:26.548+03:00 [ERR] c:[5302e508] Error during downloadOrVerifyAttachments for doc 265c50e9-7eb8-47b8-abdb-8b5a22380416/4-042718199b2b7dd97d4d332df82a008c0cd040d7: 400 Incorrect data sent for attachment with digest: sha1-gk1pJzVtmSZ6qIpxXxPd7mo6898= – db.(*blipHandler).handleRev() at blip_handler.go:763

In mobile device logs I see:

02-16 15:07:18.666 9663-9749/com.topkrabbensteam.zm.fingerobject E/SyncModule: Push-Pull Replication Error! Error code:: %d
CouchbaseLiteException{CouchbaseLite,11001,‘WebSocket connection closed by peer
(CouchbaseLite Android v2.8.0-273 (EE/release, Commit/d0dcb09fac@a10b5970562a Core/2.8.0 (273) at 2020-09-23T22:10:05.765Z) on Java; Android 6.0; X5max_PRO)’}
at com.couchbase.lite.internal.CBLStatus.toCouchbaseLiteException(CBLStatus.java:96)
at com.couchbase.lite.internal.CBLStatus.toCouchbaseLiteException(CBLStatus.java:58)
at com.couchbase.lite.AbstractReplicator.updateStatus(AbstractReplicator.java:858)
at com.couchbase.lite.AbstractReplicator.c4StatusChanged(AbstractReplicator.java:750)
at com.couchbase.lite.AbstractReplicator$ReplicatorListener.lambda$statusChanged$0(AbstractReplicator.java:254)
at com.couchbase.lite.-$$Lambda$AbstractReplicator$ReplicatorListener$VQV3yukqXlbZQTDkVEhKdctJGcw.run(lambda)
at com.couchbase.lite.internal.AbstractExecutionService$InstrumentedTask.run(AbstractExecutionService.java:94)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1113)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:588)
at java.lang.Thread.run(Thread.java:818)

And also I see alot of these messages:

02-16 14:52:17.899 8941-9097/com.topkrabbensteam.zm.fingerobject D/NativeCrypto: ssl=0xa8e74fc0 SSL_connect:3WCH_A SSLv3 write client hello A
02-16 14:52:17.899 8941-9098/com.topkrabbensteam.zm.fingerobject D/libc-netbsd: [getaddrinfo]: hostname=msinc.ru; servname=(null); netid=0; mark=0
02-16 14:52:17.899 8941-9097/com.topkrabbensteam.zm.fingerobject D/NativeCrypto: ssl=0xa8e74fc0 info_callback ignored
02-16 14:52:17.899 8941-9097/com.topkrabbensteam.zm.fingerobject D/NativeCrypto: ssl=0xa8e74fc0 info_callback where=0x1002 ret=-1
02-16 14:52:17.899 8941-9098/com.topkrabbensteam.zm.fingerobject D/libc-netbsd: [getaddrinfo]: ai_addrlen=0; ai_canonname=(null); ai_flags=4; ai_family=0
02-16 14:52:17.899 8941-9097/com.topkrabbensteam.zm.fingerobject D/NativeCrypto: ssl=0xa8e74fc0 SSL_connect:error exit in 3RSH_A SSLv3 read server hello A
02-16 14:52:17.899 8941-9097/com.topkrabbensteam.zm.fingerobject D/NativeCrypto: ssl=0xa8e74fc0 info_callback ignored
02-16 14:52:17.899 8941-9097/com.topkrabbensteam.zm.fingerobject D/NativeCrypto: doing handshake – ret=-1
02-16 14:52:17.899 8941-9097/com.topkrabbensteam.zm.fingerobject D/NativeCrypto: ssl=0xa8e74fc0 NativeCrypto_SSL_do_handshake ret=-1 errno=11 sslError=2 timeout_millis=30000

Is this some sort of a SSL problem or what?

UPDATE: Found out that it happens on some devices when amount of attachements quite big like 300 documents with 300 attachments. If I make like 20 docs with 20 attachments then everything seems to be fine.

Does SG or CBL have some replication size limitations for some devices?

Duplicate of CBL 2.8 Attachment errors "400 Incorrect data sent for attachment with digest" - #6 by Nikolay_Nikolaev. Will close.