Twice I have gotten the same error:

2018-11-20 15:43:41 : You need to rebuild the database using -reindex to change -txindex. Please restart with -reindex or -reindex-chainstate to recover. 2018-11-20 15:43:41 Aborted block database rebuild. Exiting. 

Here is my procedure that results in the error:

• Shutdown bitcoind on server #1
• ran rsync -Lru --info=progress2 /path/to/blockchain /path/to/external/usb (note: /path/to/external/usb is a BLANK directory on the external drive)
• unmounted external USB
• plug into Raspberry Pi, boot
• Made sure blockchain was visible in /home/USER/.bitcoin
• Ran bitcoind at the terminal.

Notes:

• txindex is turned on in bitcoin.conf on server #1
• txindex is also on in bitcoin.conf in /home/user/.bitcoin. on the raspberry pi I can verify bitcoind on the raspberry pi is loading that exact .conf, see the log below.
• commenting out #txindex in the bitcoin.conf file and simply running bitcoind with no flags results in a successful startup.

Here's the output of my log file when it failed:

 2018-11-20 15:43:13 Bitcoin Core version v0.16.3 (release build) 2018-11-20 15:43:13 InitParameterInteraction: parameter interaction: -whitelistforcerelay=1 -> setting -whitelistrelay=1 2018-11-20 15:43:13 Assuming ancestors of block 0000000000000000005214481d2d96f898e3d5416e43359c145944a909d242e0 have valid signatures. 2018-11-20 15:43:13 Setting nMinimumChainWork=000000000000000000000000000000000000000000f91c579d57cad4bc5278cc 2018-11-20 15:43:13 Using the 'standard' SHA256 implementation 2018-11-20 15:43:14 Default data directory /home/satoshi/.bitcoin 2018-11-20 15:43:14 Using data directory /home/satoshi/.bitcoin 2018-11-20 15:43:14 Using config file /home/satoshi/.bitcoin/bitcoin.conf 2018-11-20 15:43:14 Using at most 64 automatic connections (1024 file descriptors available) 2018-11-20 15:43:14 Using 16 MiB out of 32/2 requested for signature cache, able to store 524288 elements 2018-11-20 15:43:14 Using 16 MiB out of 32/2 requested for script execution cache, able to store 524288 elements 2018-11-20 15:43:14 Using 4 threads for script verification 2018-11-20 15:43:14 scheduler thread start 2018-11-20 15:43:14 HTTP: creating work queue of depth 16 2018-11-20 15:43:14 No rpcpassword set - using random cookie authentication 2018-11-20 15:43:14 Generated RPC authentication cookie /home/satoshi/.bitcoin/.cookie 2018-11-20 15:43:14 HTTP: starting 4 worker threads 2018-11-20 15:43:14 Cache configuration: 2018-11-20 15:43:14 * Using 50.0MiB for block index database 2018-11-20 15:43:14 * Using 8.0MiB for chain state database 2018-11-20 15:43:14 * Using 342.0MiB for in-memory UTXO set (plus up to 286.1MiB of unused mempool space) 2018-11-20 15:43:14 init message: Loading block index... 2018-11-20 15:43:14 Opening LevelDB in /home/satoshi/.bitcoin/blocks/index 2018-11-20 15:43:15 Opened LevelDB successfully 2018-11-20 15:43:15 Using obfuscation key for /home/satoshi/.bitcoin/blocks/index: 0000000000000000 2018-11-20 15:43:40 LoadBlockIndexDB: last block file = 1437 2018-11-20 15:43:40 LoadBlockIndexDB: last block file info: CBlockFileInfo(blocks=57, size=71779403, heights=550735...550791, time=2018-11-19...2018-11-20) 2018-11-20 15:43:40 Checking all blk files are present... 2018-11-20 15:43:41 LoadBlockIndexDB: transaction index disabled 2018-11-20 15:43:41 : You need to rebuild the database using -reindex to change -txindex. Please restart with -reindex or -reindex-chainstate to recover. 2018-11-20 15:43:41 Aborted block database rebuild. Exiting. 2018-11-20 15:43:41 Shutdown: In progress... 2018-11-20 15:43:41 scheduler thread interrupt 2018-11-20 15:43:41 Shutdown: done 

I COULD just run bitcoind -reindex but that doesn't solve the problem of 'what is causing the txindex database to not work between machines.' I feel like this is an important question to answer.