Activity for UFTP

Sean Summers posted a comment on discussion General Discussion

Sean Summers — Fri, 19 Sep 2025 21:50:19 -0000

Thanks @dennisbush, you're totally right. I didn't include the -c option and I was expecting that behavior.

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Wed, 17 Sep 2025 17:52:38 -0000

Sean, Encryption is in fact working. There's just no verification of the remote machines' keys happening. The initial ANNOUNCE message from the server includes the server's public signing key, and the CLIENT_KEY response from the client (which requires the -c option on the server) contains the client's public signing key. The -H option on the server can specify the expected key fingerprint for each client, and similarly the -S option on the client can specify the expected key fingerprint for each...

Sean Summers posted a comment on discussion General Discussion

Sean Summers — Wed, 17 Sep 2025 04:51:23 -0000

I'm unable to get any sort of meaningful encryptiong working with UFTP. I'm running the server with a private key that uftp generated: uftp -Y aes256-gcm -h sha256 -k uftp_priv.pem -R 40000 file.txt One client like this: uftpd -d -E -k private_key.pem -D /.firmware And the other like this: uftpd -d -E -D /.firmware I haven't registered the clients' public keys with the server, nor vice versa, but both receivers still get the file. I would expect: (A) without the server having the clients' public...

Sean Summers posted a comment on discussion General Discussion

Sean Summers — Tue, 12 Aug 2025 20:26:52 -0000

Thanks for responding, I hadn't considered cross platform transfers; I'm going Linux->Linux. I suppose you could do this if uftp compressed a directory and uftpd decompressed it. For now, I think I'll have to send some metadata with the directory manually and run a script to restore permissions

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Tue, 12 Aug 2025 19:33:33 -0000

That's tricky, because how file permissions are implemented vary across operating systems. You'd likely have a similar problem if you were to copy a from from Linux to Windows back to Linux for example. Directory transfers are implemented as multiple individual file transfers, so there's no good way to communicate file permissions in a portable way without creating an archive.

Sean Summers posted a comment on discussion General Discussion

Sean Summers — Tue, 12 Aug 2025 18:19:28 -0000

It seems like uftp doesn't preserve file permissions when transmitting a directory of files. I would get around this by tar'ing up the directory first, but my receivers don't have enough space for the tar archive and the extracted directory of files. Is there a way to preserve file permissions natively?

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Fri, 02 May 2025 03:18:32 -0000

Yegor, I don't have plans to add LibreSSL support at this time as I haven't gotten any inquiries about it until now. If demand increases, I may revisit this. Regards, Dennis

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Fri, 02 May 2025 03:16:30 -0000

Yegor, Changing this code to use the EVP_PKEY function instead of the RSA and EC functions will take some time, and I'm not sure if doing so will break systems still using OpenSSL 1.1. In the meantime, you can add -Wno-deprecated-declarations to the CFLAGS variable in the makefile to silence the warnings. Regards, Dennis

Yegor Yefremov posted a comment on discussion General Discussion

Yegor Yefremov — Tue, 29 Apr 2025 08:41:22 -0000

Hi Dennis, are you planning to add support for LibreSSL? Building against it produces the following linking errors: /home/buildroot/instance-0/output-1/host/lib/gcc/powerpc64le-buildroot-linux-gnu/13.3.0/../../../../powerpc64le-buildroot-linux-gnu/bin/ld: encrypt_openssl.c:(.text+0x380): undefined reference to EVP_add_cipher' /home/buildroot/instance-0/output-1/host/lib/gcc/powerpc64le-buildroot-linux-gnu/13.3.0/../../../../powerpc64le-buildroot-linux-gnu/bin/ld: encrypt_openssl.o: in functioncrypto_init':...

Yegor Yefremov posted a comment on discussion General Discussion

Yegor Yefremov — Mon, 28 Apr 2025 13:30:23 -0000

Hi Dennis, I get a bunch of -Wdeprecated-declarations warnings on GCC 14.2.0: encrypt_openssl.c: In function ‘RSA_keylen’: encrypt_openssl.c:487:5: warning: ‘RSA_size’ is deprecated: Since OpenSSL 3.0 [-Wdeprecated-declarations] 487 | return RSA_size(rsa); | ^~~~~~ In file included from encrypt_openssl.c:43: /usr/include/openssl/rsa.h:215:27: note: declared here 215 | OSSL_DEPRECATEDIN_3_0 int RSA_size(const RSA *rsa); | ^~~~~~~~ encrypt_openssl.c: In function ‘EC_keylen’: encrypt_openssl.c:501:5:...

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Fri, 25 Apr 2025 21:54:39 -0000

Hi Dave, you can send them to my email.

Dave MacLachlan posted a comment on discussion General Discussion

Dave MacLachlan — Fri, 25 Apr 2025 19:34:50 -0000

Hey Dennis.. I have some patches to fix up some macOS issues. How is best to send them to you?

Dave MacLachlan posted a comment on discussion General Discussion

Dave MacLachlan — Fri, 25 Apr 2025 19:34:09 -0000

This appears to be specific to one of the network I am running on so likely not a UFTP issue.

Dave MacLachlan posted a comment on discussion General Discussion

Dave MacLachlan — Fri, 25 Apr 2025 18:22:36 -0000

Thank you for uftp. Built 5.0.3 for macOS and linux. I have noticed on both macos and linux (ubuntu) that it appears that uftpd stops receiving packets after a couple of minutes after launch. I am working to get more details, but wondering if I am holding it wrong? uftpd -d -D /Users/dmaclach/Desktop nothing interesting appearing with -x 5

UFTP updated /Changes.txt

UFTP — Mon, 18 Dec 2023 02:54:02 -0000

UFTP released /source-tar/uftp-5.0.3.tar.gz

UFTP — Mon, 18 Dec 2023 02:54:02 -0000

UFTP released /source-zip/uftp_src-5.0.3.zip

UFTP — Mon, 18 Dec 2023 02:54:02 -0000

UFTP released /exe-windows/uftp_exe_W7_x64-5.0.3.zip

UFTP — Mon, 18 Dec 2023 02:54:02 -0000

Eriza Fazli posted a comment on discussion General Discussion

Eriza Fazli — Tue, 28 Nov 2023 21:09:47 -0000

Thanks Dennis!

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Tue, 28 Nov 2023 21:05:44 -0000

Yes, this was a bug introduced in the most recent version. There was an update to fix a memory leak, but the fix wasn't applied correctly. I'll send out an updated release soon.

Eriza Fazli posted a comment on discussion General Discussion

Eriza Fazli — Tue, 28 Nov 2023 19:14:45 -0000

Hi, client crashes when using server options -Y none -C tfmcc. Both client and server are compiled with encryption. I tried swapping client and server (raspberry pi and laptop) with the same outcome. I was wondering if I might be missing something? Here below is a snippet: server: ./uftp -R 600000 -B 20971520 -f -x 5 -Y none -C tfmcc:min=20000:init=600000 1GB.bin client: ./uftpd -d -B 20971520 -c 20971520 -x 5 client output: 2023/11/28 18:48:02.809008: UFTP version 5.0.2 Copyright (C) 2001-2023 Dennis...

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Thu, 09 Nov 2023 01:45:18 -0000

Vit, Thanks for catching this. I've just released version 5.0.2 to address this issue as well as some memory leaks recently uncovered. Regards, Dennis

UFTP released /source-tar/uftp-5.0.2.tar.gz

UFTP — Thu, 09 Nov 2023 01:42:03 -0000

UFTP released /source-zip/uftp_src-5.0.2.zip

UFTP — Thu, 09 Nov 2023 01:42:03 -0000

UFTP released /exe-windows/uftp_exe_W7_x64-5.0.2.zip

UFTP — Thu, 09 Nov 2023 01:42:03 -0000

UFTP updated /Changes.txt

UFTP — Thu, 09 Nov 2023 01:42:03 -0000

jtnf posted a comment on discussion General Discussion

jtnf — Mon, 06 Nov 2023 18:37:16 -0000

I've been very happy with uftpd, solving some particularly sticky problems in a distributed network. I'd like to extend my testing to Android devices, and was wondering if anyone had any pointers on compiling for Andriod 10 or 11. I have much experience compiling and supporting OSS on Linux and all the commercial UNIXen, but have never done any ports to Android. I have the Android IDE recently installed on my Windows laptop, and I'm about to dive into researching building services and daemons.

Vít Lapka posted a comment on discussion General Discussion

Vít Lapka — Thu, 19 Oct 2023 12:06:24 -0000

Hello, while testing with multiple clients and on a more congested network, we have run into crashes of UFTP server in Announce phase (with a closed group). This was caused by REGISTER packet from client being lost in the network, but CLIENT_KEY packet arriving. In that case the server attempted to decrypt CLIENT_KEY but had no encryption context allocated (normally done on receiving REGISTER), leading to the crash (a memory violation). I have made a fix by adding a check in server_phase.c:handle_announce_phase()...

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Fri, 22 Sep 2023 02:25:22 -0000

René, Thanks for catching this. I'll include it in the next release. Regards, Dennis

René Samek posted a comment on discussion General Discussion

René Samek — Thu, 21 Sep 2023 08:06:02 -0000

Hello, I found a memory leak using UFTP with TFMCC congestion control. Function send_cc_ack(...) ("client_transfer.c" module) does't free encrypted buffer when encrypt_and_sign(...) call is successfull. On top of that I found a potential memory leak in the function encrypt_and_sign() ("uftp_common.c" module). In the attachment is the patch I used to fix the problems. With regards René Samek

Robert Bui posted a comment on discussion General Discussion

Robert Bui — Mon, 06 Feb 2023 21:56:10 -0000

Looking at the high System Interrupts, I should have investigated the cause of that. Also, setting the UFTP log level (-x flag) on the client side to 4, I've found the client missed a whole section at regular interval. So, most likely the network is the issue. Downloaded the latest version of the network card driver (12 months old) and upgraded from the current version (4 years old), immediately saw a huge improvement. No more NAKs in the 10, 000s. I am seeing a lot of NAKs < 100, with occassionly...

Robert Bui posted a comment on discussion General Discussion

Robert Bui — Sun, 05 Feb 2023 21:48:12 -0000

Strange that on the computers that have high NAKs, there are two processes accessing the download file (see attached): one process is uftp.exe and the other process is System. Both constantly appearing on the Resources Monitor doing disk access. While on the computers with low NAKs, onlythe System process accessing the download file. Maybe there is an uftp.exe process but it is so fast that not appearing or appear for a very short time ont eh Resources Monitor screen.

Robert Bui posted a comment on discussion General Discussion

Robert Bui — Wed, 01 Feb 2023 02:09:26 -0000

Hello all, After we've upgrade our computers from Windows 10 version 1903 to version 21H2, we are getting excessive NAKs, when transfer >100GB data file across Gigabit local network. On 1903, we may receive on average 200-400 NAKs per failed section. We may get a failed section every 100-200 sections. It took 20mins to transfer. After we've upgraded to version 21H2 (even happening on version 22H2), we are getting on average 1000-2000 NAKs per failed setion. We get a failed section every 10-15 sections....

Vít Lapka posted a comment on discussion General Discussion

Vít Lapka — Thu, 29 Sep 2022 08:40:18 -0000

Hello, I have been testing UFTP with TFMCC on low latency link (RTT ~ 27ms) with severely limited maximal rate (down to 1Kb/s): -C "tfmcc:init=1:max=1". However, during the transfer of test file the rate climbed above the limit. This is caused by order of conditions in function server_transfer.c:handle_tfmcc_ack_info() that updates the rate based on CC_ACK data: max_rate is checked and enfoced min_rate is checked and enforced minimal rate of 1 packet per RTT is enforced The last condition may violate...

UFTP released /exe-windows/uftp_exe_W7_x64-5.0.1.zip

UFTP — Wed, 03 Aug 2022 01:28:03 -0000

UFTP released /source-zip/uftp_src-5.0.1.zip

UFTP — Wed, 03 Aug 2022 01:28:03 -0000

UFTP released /source-tar/uftp-5.0.1.tar.gz

UFTP — Wed, 03 Aug 2022 01:28:03 -0000

UFTP updated /Changes.txt

UFTP — Wed, 03 Aug 2022 01:28:03 -0000

Vít Lapka posted a comment on discussion General Discussion

Vít Lapka — Fri, 29 Jul 2022 11:58:48 -0000

Hello, I am sending another patch, but it is not a fix, it adds a new functionality. I do not propose to add it to the project, I am sending it only in case you find it interesting. Since I will be running some long-term UFTP processes, I need a feedback. So I have devised a new parameter, for both the client and the server (I am not using a proxy at this moment): '-G progfile'. It allows to set up a file for progress reporting. This file can be the logging stream ("@LOG"). When the process receives...

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Wed, 27 Jul 2022 23:11:19 -0000

Yes, that is a bug. The comparisons should each be == instead of !=. I'll fix that as well. The conditions that follow those lines are correct.

Vít Lapka posted a comment on discussion General Discussion

Vít Lapka — Tue, 26 Jul 2022 09:11:56 -0000

Hello, I have noticed a suspicious condition in client_fileinfo.c: handle_fileinfo():539: if ((group->fileinfo.ftype != FTYPE_DELETE) || (group->fileinfo.ftype != FTYPE_FREESPACE)) { This condition is always fulfilled, which disables following code dealing with moving an existing target file to backup. There are similarly looking conditions in this section, that are now never evaluated. I have not attempted to do anything with this, since it requires further analysis and it lays outside of my use...

Vít Lapka posted a comment on discussion General Discussion

Vít Lapka — Mon, 18 Jul 2022 06:01:21 -0000

Thank you for the reply. Fortunately my use case was simplified to using only one interface and it already has a script called whenever it is created or destroyed and it should be a rare event. Regarding GRTT: If I understand it correctly, then from the standpoint of client it represents the expected time of arrival of the next packet. Various timeouts are derived from this. If the timeout runs from reception of one packet to reception of another, it should be derived from sending delay + 1/2 * RTT....

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Sun, 17 Jul 2022 01:47:42 -0000

Vit, Thanks for the patches. Most of these I can include as is (with minor formatting changes). The changes to the advertised GRTT will need a little refactoring. One thing in particular I might do with that is to make the inter-packet wait time the lower limit on the advertised GRTT rather than just adding to it. That should still be enough for your use case while preventing any potential adverse affects on higher transfer speeds. As for the issue of interfaces going up and down, there's probably...

Vít Lapka posted a comment on discussion General Discussion

Vít Lapka — Fri, 15 Jul 2022 12:22:51 -0000

One more thing I have run into and have not attempted to solve yet: If the client is configured to listen on a dynamic interface that can get removed (not just set DOWN) and created again, it will stop listening (loose its registration for multicast reception). From the standpoint of the list of interfaces, it is probably an entirely new interface with possibly a new index. The only solution that worked so far was restarting the client and then restarting the transaction from the server. I guess...

Vít Lapka posted a comment on discussion General Discussion

Vít Lapka — Fri, 15 Jul 2022 12:20:38 -0000

Hello, I have been testing UFTP 5.0 on the embedded Linux with OpenSSL 1.1.1o. I was trying to use it over a narrowband wireless link. The idea is to have it run in the background with a severely limited maximal rate (as low as 1Kbps) and to rely on session restart/resume if endpoints get disconnected or rebooted. The results are looking promising, thank you for creating UFTP. I have run into several issues and created patches to address them. I am sending them in the attachments in case you would...

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Tue, 26 Apr 2022 21:20:50 -0000

Start by running Wireshark on both the sending and receiving machine to see if the packets are appearing on the local interfaces.

Maxim Pavlov posted a comment on discussion General Discussion

Maxim Pavlov — Sat, 23 Apr 2022 08:15:29 -0000

Hello, I an new to UFTP, and I wish to just copy files via multicast from IP 192.168.1.2 to 192.168.1.10. On the PC with 192.168.1.10 I give the command uftpd.exe -I 192.168.1.10 -L c:\temp\log.txt On the PC with 192.168.1.2 I give the command uftp.exe -I 192.168.1.2 file_to_transfer UFTP version 5.0 Copyright (C) 2001-2020 Dennis A. Bush Starting at Sat Apr 23 14:30:33 2022 Loaded ECDSA key with curve prime256v1 and fingerprint A6:E0:54:A6:EC:28:33:C4:EA:89:06:00:82:0F:16:62:38:8A:B3:AE Loaded ECDH...

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Tue, 24 Aug 2021 21:30:53 -0000

Yes, the change to the -r option sets the minimum round trip to 0.5 seconds, and adjusting that value (specifically the second one) will help reduce the overall time.

jason znack posted a comment on discussion General Discussion

jason znack — Tue, 24 Aug 2021 17:35:29 -0000

WellI can confirm now that the transfers are successfully completing. However the average time has gone from 20 seconds to 60 seconds. If I understand what the RTT change has done, is that the minimum time per file is now 0.5 seconds round trip instead of 0.01 seconds? And so tuning that minimum time to find the point where transfers start to fail again might allow for a reduction in transfer times again? I think getting rid of that problem client entirely will be the long term solution. Or at least...

jason znack posted a comment on discussion General Discussion

jason znack — Tue, 24 Aug 2021 05:17:02 -0000

It seems a bit better with the AV and priority changes alone wihtout the RTT changes, but it still drops the problem client eventually. 2021/08/24 00:58:47.395362: [E66DD6D4/00:0]: ----- DailyVideos/G2021-08-23-2202-A/2021-08-24-0056-A/Pictures/2_002.jpg ----- 2021/08/24 00:58:47.395362: [E66DD6D4/00:0195]: File ID: 0195 Name: DailyVideos/G2021-08-23-2202-A/2021-08-24-0056-A/Pictures/2_002.jpg 2021/08/24 00:58:47.395362: [E66DD6D4/00:0195]: sending as: DailyVideos/G2021-08-23-2202-A/2021-08-24-0056-A/Pictures/2_002.jpg...

jason znack posted a comment on discussion General Discussion

jason znack — Tue, 24 Aug 2021 04:54:24 -0000

The only AV is the windows malware protection. I've setup an exclusion for the uftpd.exe process and the entire external drive that it writes to. Even when I temporarily turn it off it didn't make a difference. Plus it always turns itself back on anyway. I also added dropbox.exe to the exclusion list since it's the only other process that is a resource hog. I've increased the priority to -N -2. Are you sure you mean -g for RTT? The manual that I've been looking at use that for max log size. -g max_log_size...

Dennis Bush modified a comment on discussion General Discussion

Dennis Bush — Mon, 23 Aug 2021 20:29:16 -0000

Regarding the "incorrect file_id" message, each file in a session is assigned a sequential number starting with 1. This message is printed when a message comes in related to one file which is different from the active file. It's not uncommon to see this message when the invalid file_id is for the prior file, but the fact that the file_id you're seeing is far off from the current file and in some cases larger is concerning. I'm almost positive that the disk spikes are coming from some unrelated process...

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Mon, 23 Aug 2021 20:23:15 -0000

jason znack posted a comment on discussion General Discussion

jason znack — Mon, 23 Aug 2021 06:43:41 -0000

Rejecting FILESEG: got incorrect file_id 0169 what exactly is this message telling me? What file is file_id 0169? It's usually 0169 in the logs, but it's sometimes others as well like 0139 or 014D or 014E or 014F or 0150. Sometimes it happens once, sometimes it happens on hundreds of lines repeating. I tried -N -1 with no real change. Tried different transfer rates, buffer sizes, cache size, congestion control, etc. I thought I had found a drive latency problem with the external hard drive taking...

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Fri, 20 Aug 2021 12:44:20 -0000

This definitely looks like disk latency. During that 4 second gap in the client log, the only thing the client is doing is checking the validity of the destination filename, creating directories in the target file's path if required, and reading the info for the existing file. Nothing resource intensive. See if there's anything else that might be using that disk. Antivirus could be a possibility. Try temporarily disabling it add see what happens. Also, increasing the client's priority might give...

jason znack posted a comment on discussion General Discussion

jason znack — Fri, 20 Aug 2021 07:31:37 -0000

I'll see if I can grab more incorrect file ID messages. I have seen a lot of those. 2021/08/19 21:31:30.161130: [788D22D6/00:0151]: Name of file to receive: DailyVideos/G2021-08-19-2106-A/2021-08-19-2117-A/Pictures/2_000.jpg 2021/08/19 21:31:30.161130: [788D22D6/00:0151]: Bytes: 328462, Blocks: 41, Sections: 1 2021/08/19 21:31:30.161130: [788D22D6/00:0151]: copying new file 2021/08/19 21:31:30.161130: [788D22D6/00:0151]: FILEINFO_ACK sent 2021/08/19 21:31:30.176755: [788D22D6/00:0151]: Got DONE message...

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Fri, 20 Aug 2021 03:24:46 -0000

Jason, "read timeout" is a normal message at log level 5 that states how long the client will wait for a packet before checking if any message related timers have tripped in order to send retransmissions or to check if the user pressed CTRL-C. It's not indicative of a problem. The incorrect file ID messages are interesting. Unless there is another session going on that shouldn't happen. The NAKs in the log seem to indicate that the client is loosing packets in bursts, possibly because some other...

jason znack modified a comment on discussion General Discussion

jason znack — Fri, 20 Aug 2021 00:12:53 -0000

I have one client that refuses to get any files and I can't figure out why. Any help or insight would be greatly appreciated. Client command: C:\uftp\uftpd.exe -L uftplog.log -x 5 -g 10 -n 100 -c 10485760 -B 104857600 -D D:\ Sender command: C:\UFTP\uftp.exe -R -1 -S status.txt -L uftplog.log -g 10 -n 100 -Y none -m 10 -b 8192 -B 2097152 -z -s 50 C:\DailyVideos It's syncing a folder of photos and videos to 7 windows pcs. Usually it's just a single new folder that gets added inside of that root folder....

jason znack modified a comment on discussion General Discussion

jason znack — Thu, 19 Aug 2021 21:55:47 -0000

jason znack modified a comment on discussion General Discussion

jason znack — Thu, 19 Aug 2021 21:53:14 -0000

jason znack posted a comment on discussion General Discussion

jason znack — Thu, 19 Aug 2021 21:47:06 -0000

jason znack posted a comment on discussion General Discussion

jason znack — Mon, 05 Jul 2021 22:11:57 -0000

Thanks, I'll do that if it happens again.

jason znack posted a comment on discussion General Discussion

jason znack — Mon, 05 Jul 2021 22:10:29 -0000

Thanks. That jives with what I was able to see from some testing (Edited my post above with results) I'll do some further testing to fine tune.

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Mon, 05 Jul 2021 22:08:45 -0000

When troubleshooting issues like this, try doing a packet capture on both sides to make sure the network adapter is actually receiving the data.

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Mon, 05 Jul 2021 22:07:23 -0000

I'd say go with at least 1MB (1048576) for the UDP buffer size. You can also adjust that up or down as needed.

jason znack modified a comment on discussion General Discussion

jason znack — Mon, 05 Jul 2021 22:06:32 -0000

How about the UDP buffer size? Leave as default? After playing with the cache it seems to be best around 3mb, and increasing the UDP buffer to 1500k seems to be better. I've applied the FastSendDatagramThreshold 1500 registry fix for windows. Preliminary results from today's testing: -c 2097152 -B 153600 = 90 sec -c 2097152 -B 768000 = 22 sec -c 2097152 -B 1536000 = 20 sec -c 2097152 -B 3072000 = 26 sec -c 3145728 -B 1536000 = 25 sec -c 10485760 -B 104857600 = 34 sec

jason znack posted a comment on discussion General Discussion

jason znack — Mon, 05 Jul 2021 21:39:58 -0000

And today I try again and it appears to be working ok. Nothing changed though as far as I know.

jason znack posted a comment on discussion General Discussion

jason znack — Mon, 05 Jul 2021 19:38:17 -0000

How about the UDP buffer size? Leave as default?

jason znack posted a comment on discussion General Discussion

jason znack — Mon, 05 Jul 2021 19:24:19 -0000

Hello, I'm having some trouble getting a transfer to start on a machine that is using a 2x1Gb Link Aggregation The NIC is an Intel Dual gigabit and the LAG is setup as 802.3ad to a Unifi switch. Normal transfers work fine with activity on both links. However, when I try to initiate a UFTP transfer, it announces, and then fails with Sending thread timed out and Failed to join sender thread. I can successfully send transfers from another machine on the network that doesn't have a dual port NIC, and...

jason znack posted a comment on discussion General Discussion

jason znack — Sun, 04 Jul 2021 06:36:23 -0000

Thanks a lot. I'll give that a try. On Sat, Jul 3, 2021 at 10:07 PM Dennis Bush dennisbush@users.sourceforge.net wrote: Jason, Try setting the block size to 8192, that way it's a multiple of the disk allocation size. Playing with the cache size should also have an effect. First try 2097152 (2MB) then go up in increments of 1048576 (1MB) to see how that changes the throughput. Regards, Dennis Tuning Advice https://sourceforge.net/p/uftp-multicast/discussion/general/thread/d4b178fbca/?limit=25#9fc4...

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Sun, 04 Jul 2021 04:07:26 -0000

Jason, Try setting the block size to 8192, that way it's a multiple of the disk allocation size. Playing with the cache size should also have an effect. First try 2097152 (2MB) then go up in increments of 1048576 (1MB) to see how that changes the throughput. Regards, Dennis

jason znack posted a comment on discussion General Discussion

jason znack — Sat, 03 Jul 2021 18:33:18 -0000

Hello, I'm trying to understand how best to tune the cache and buffer parameters for best performance. The setup is as follows 1 server, and 6 clients on a gigabit network. All systems are pretty low end, but are running Windows 10 and have SSDs. Using version 5. No encryption. Unlimited rate. The data being sent is a couple larger 100mb or so sized files and a dozen smaller 1mb sized files. Jumbo frames are enabled on the network and all clients, so the block size is 8800. Using the defaults for...

tobias frahm posted a comment on discussion General Discussion

tobias frahm — Thu, 24 Sep 2020 06:14:51 -0000

Hello Dennis, the high amount of NAKs was due to the client beeing busy. Since I could not observe this, I thought about the Storage Speed. I already knew this was a bottleneck but I expected this to work. Since the SD-Card has a write speed of about 15MB/s (120000Kbps) but I plugged in a USB flashdrive anyways and it is way more stable with the flashdrive instead of the sd-cad. INFO:root:Total NAKs found 1469.0 INFO:root:Avr throughput without fails: 34.26 Mbps INFO:root:Avr time needed without...

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Wed, 23 Sep 2020 14:09:35 -0000

Tobi, The -C option is used to enable congestion control. The algorithm it uses (TFMCC, RFC4654) does tend to be quick to scale back the speed in the face of a slowdown. You could try changing the blocksize (-b) on the server to 1024 to better fit the disk's block size. Also, try varying the write cache size on the client (-c) to see what works. Getting the right value for -c will probably do a lot for your throughput. Regards, Dennis

tobias frahm posted a comment on discussion General Discussion

tobias frahm — Wed, 23 Sep 2020 08:55:54 -0000

Hello Dennis, thank you for your reply. So the -Y none was a good shot. Since I already figured out that encryption is indeed a bottleneck on the raspberry. But if it changed something, its not really recognizeable. I logged the CPU load during the transfer on raspberry site. -Y none, the maximum CPU load was about 10% in this case Client: pi@raspberrypi:~ $ uftpd -d -D ~/Downloads 2020/09/23 10:14:16.984219: UFTP version 5.0 Copyright (C) 2001-2020 Dennis A. Bush 2020/09/23 10:14:16.986478: Loaded...

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Tue, 22 Sep 2020 14:17:51 -0000

Tobias, You're getting a very high level of NAKs. Based on the total number of blocks and sections, each section has either 10324 or 10325 blocks, and there are several sections which show NAKs for all blocks in several consecutive sections. One thing in particular that jumped out from the log was this: 2020/09/22 10:27:23.914010: [15990A7E/00:0001]: Sent 7519 NAKs for section 14 2020/09/22 10:27:44.748964: [15990A7E/00:0001]: Sent 1279 NAKs for section 16 2020/09/22 10:27:44.751497: [15990A7E/00:0001]:...

tobias frahm modified a comment on discussion General Discussion

tobias frahm — Tue, 22 Sep 2020 08:58:31 -0000

Hello, I am facing a unreliable transfer. Setup: I have a performant Laptop (1Gbps ethernet interface) with manjaro Linux running. My clients are Raspberry 3B+ (no quite sure whats possible, but I read about 100Mbps up to 340 Mbps) I target less than 100Mbps, less is Okay, but I need it stable. I am Running Tests, sening a 1GB file from one server to two clients. Created Custom logs, just because I am too lazy to go through more than 300 Logs and python did sum-up the work. One bug to mention: Total...

tobias frahm posted a comment on discussion General Discussion

tobias frahm — Tue, 22 Sep 2020 08:57:11 -0000

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Thu, 03 Sep 2020 21:28:53 -0000

This indicates that a client aborted the session because no messages had been received from the server after some time. Exactly how long depends on the current GRTT (group round trip time) and which phase of the session the client is in. Try running a Wireshark trace on the client to see exactly what is coming in and when.

Kacper Marczyk posted a comment on discussion General Discussion

Kacper Marczyk — Thu, 03 Sep 2020 14:11:01 -0000

Hi again Dennis, when we are sending only one session problem occurs less times then before. But sometimes we still gets such error: Transfer aborted by 0xF58BD285: Transfer timed out Connection works fine on both of devices. Do you have any information what this error stands for?

restlessmindsstudio posted a comment on discussion General Discussion

restlessmindsstudio — Wed, 08 Jul 2020 15:23:45 -0000

Alright after I upgraded the clients to 5.0 it's working now and CPU usage is around 5% on the server. Still not sure what changed to cause this problem in the first place. Thanks again, Jim

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Wed, 01 Jul 2020 15:07:12 -0000

It's the new client that's compatible with the old server, not the other way around. You would need to upgrade the clients to 5.0 before upgrading the server to 5.0.

restlessmindsstudio posted a comment on discussion General Discussion

restlessmindsstudio — Wed, 01 Jul 2020 05:43:45 -0000

Hi Dennis, I upgraded to 4.10.2 first and had the same issue so I then went to 5.0. With 5.0 it terminates with an exit code of 7 which I see is due to no response from clients. Clients are running uftp 4.10 and based on what I read in the release notes, it should be backward compatible. I also tried including the option: -Y none when starting the server since I don't use encryption but that still resulted in exit code 7. Looks like I need to use gdb at this point. Jim

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Tue, 30 Jun 2020 14:00:05 -0000

Jim, The version you're running is over 5 years old. I'd first recommend upgrading to at least 4.10.2, if not 5.0. If you still see the issue after upgrading, try attaching to the running process with gdb the next time you see this happening. That should give you some idea of where it's getting hung up (you'll probably need a copy of the source code on the machine for gdb to read from). Start by looking at the variables that control the round trip time. Regards, Dennis

François Corvaisier posted a comment on discussion General Discussion

François Corvaisier — Tue, 30 Jun 2020 06:30:46 -0000

It's working, thank you for your quick respons time.

restlessmindsstudio posted a comment on discussion General Discussion

restlessmindsstudio — Tue, 30 Jun 2020 05:43:27 -0000

I've been using uftp 4.5.1 on Ubuntu 16.04 64-bit at over 50 remote sites and it's been solid. Over the weekend one of the locations reported an issue with the machine hosting uftp becoming unresponsive. After some investigation I found uftp was using 100% of the CPU time. I terminated uftp and reviewed the logs but didn't see anything unusual from it. I started the process again and sure enough it pegged the CPU at 100%. I use uftp to distribute video files to other clients on a LAN. File sizes...

François Corvaisier posted a comment on discussion General Discussion

François Corvaisier — Mon, 29 Jun 2020 17:23:51 -0000

Oh no... I just look at the wrong help file... it's for the version 1 :'( I'll test tomorrow morning to check... I guess it will be good. Thanks.

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Mon, 29 Jun 2020 17:19:49 -0000

François, I believe your serverlist file is not formatted correctly. The format of the file is: server_id|server_ip|proxy_id|server_fingerprint It looks like you left out the proxy_id field, meaning that the server fingerprint was read as the proxy ID. If you're not using a proxy, this field should be 0, i.e.: 0x11111111|Server IP|0|XX:XX:XX:.... Give that a try. Regards, Dennis

François Corvaisier posted a comment on discussion General Discussion

François Corvaisier — Mon, 29 Jun 2020 16:05:48 -0000

Hi, I was looking for your tool to transfer a large amount of data, 12T, over Wan... I know it will talke some times, but I need to see if it will only work. ;) But I need to secure at different levels the transfert so I need to use certificats and serverlist restrictions, added to firewall rules. But My question here is more about the serverlist file. I have the version 5.0 and both server and client are on Windows 2019. The firewalls are off during my internal tests. My file is. : 0x11111111|Server...

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Wed, 24 Jun 2020 12:08:19 -0000

The server log is incomplete, so it's hard to see what might have happened. Also, based on the timestamps it doesn't look like the server log and client log correspond with each other. I'll need to see a complete client log and server log from the same session (i.e. the same group ID appears in both logs) to have any idea what might be happening.

Kacper Marczyk posted a comment on discussion General Discussion

Kacper Marczyk — Wed, 24 Jun 2020 09:52:23 -0000

Hi, In attachment you can find logs from server and from one of the clients. As you can see there is nothing about files from server in the client (take a look on the time of sending from server and receiving time on client).

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Mon, 22 Jun 2020 13:29:53 -0000

The "Resource temporarily unavailable" is some sort of local issue reading files. That shouldn't affect the session dropping out, however. The server is sending a FILEINFO and not getting a FILEINFO_ACK back from the client. Seeing both the server log and the associated client log when this happens will help figure out what is happening.

Kacper Marczyk modified a comment on discussion General Discussion

Kacper Marczyk — Mon, 22 Jun 2020 11:19:25 -0000

Actually logs on clients are clear - no mentions about these files. On server it still saying "Resource temporarily unavailable\nFailed to read directory". We are sending files from CIFS share in multiple sessions of UFTP server to multiple UFTPDs at the same moment. Do you think that might be the problem?

Kacper Marczyk posted a comment on discussion General Discussion

Kacper Marczyk — Mon, 22 Jun 2020 11:19:17 -0000

Actuallly logs on clients are clear - no mentions about these files. On server it still saying "Resource temporarily unavailable\nFailed to read directory". We are sending files from CIFS share in multiple sessions of UFTP server to multiple UFTPDs at the same moment. Do you think that might be the problem?

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Fri, 19 Jun 2020 21:28:19 -0000

This server log indicates that the client is not responding to a FILEINFO. What is in the client log in this case?

Kacper Marczyk modified a comment on discussion General Discussion

Kacper Marczyk — Fri, 19 Jun 2020 13:08:41 -0000

Hi, I am sending some files from UFTP server to few UFTPD in my LAN. I am running UFTP in multiple sessions. But sometimes when one file is sent to more than one localization happens something like in the attached log. Do you have any idea why it might keep disconnecting? I've changed names of directories and files. Thanks in advance for your help

Kacper Marczyk posted a comment on discussion General Discussion

Kacper Marczyk — Fri, 19 Jun 2020 13:06:15 -0000

Hi, I am sending some files from UFTP server to few UFTPD in my LAN. But sometimes when one file is sent to more than one localization happens something like in the attached log. Do you have any idea why it might keep disconnecting? I've changed names of directories and files. Thanks in advance for your help

Zip Ping modified a comment on discussion General Discussion

Zip Ping — Wed, 03 Jun 2020 19:53:49 -0000

Hi Dennis, Thanks for the speedy reply. I'm using macos mojave and ubuntu 18.04.4 (having tried each as both sender and receiver) and uftp 5.0. It turns out my issue was that I had compiled with NO_ENCRYPTION to keep it simple, but when trying again after recompiling with openssl 1.1.1g the error disappears and my transfers succeed. I'm also using -M flag on uftp to avoid multicast entirely, and this seems to be working well - e.g. uftpd -d -I and uftp -H -M...

Zip Ping posted a comment on discussion General Discussion

Zip Ping — Wed, 03 Jun 2020 19:52:49 -0000

Hi Dennis, Thanks for the speedy reply. I'm using macos mojave and ubuntu 18.04.4 (having tried each as both sender and receiver). It turns out my issue was that I had compiled with NO_ENCRYPTION to keep it simple, but when trying again after recompiling with openssl 1.1.1g the error disappears and my transfers succeed. I'm also using -M flag on uftp to avoid multicast entirely, and this seems to be working well - e.g. uftpd -d -I and uftp -H -M ...

Dennis Bush posted a comment on discussion General Discussion

Dennis Bush — Wed, 03 Jun 2020 15:17:59 -0000

Philip, What you're doing should work. What version of UFTP are you using, and on which OS? Regards, Dennis