Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | [SERVER] Use monotonic clock for measuring time | Simon Rettberg | 2017-10-19 | 10 | -74/+264 |
| | | | | | Introduces new shared source unit timing.[ch] Closes #3214 | ||||
* | [SERVER] More tweaks to alt-server selection wrt proxy cycles | Simon Rettberg | 2017-10-18 | 1 | -4/+9 |
| | | | | | This needs an overhaul; status needs to be moved from global alt-servers list to individual uplinks | ||||
* | [SERVER] Don't try to load metadata files as images; fix clang-analyzer ↵ | Simon Rettberg | 2017-10-18 | 2 | -15/+36 |
| | | | | false positives | ||||
* | [SERVER] Try to connect to different server if proxy cycle is detected | Simon Rettberg | 2017-10-17 | 4 | -9/+34 |
| | |||||
* | [*] Support hop-counting in request header, protocol version 3 | Simon Rettberg | 2017-10-17 | 10 | -45/+101 |
| | | | | | | | | | | | | | | We steal 8 bits from the request offset to count hops when requests get relayed by proxies. This still leaves plenty of bits for the offset (56 bits, supporting images of up to 72 petabytes). This is used to detect proxy cycles. The algorithm is not perfect but should prevent endless relays of the same request. This is backwards compatible to old clients and servers, as the server only ever sets the hopcount in relayed requests if the upstream server is using protocol version 3 or newer, and clients are automatically upwards compatible as there is practically no image larger than 74PB, so the newly introduced hop count field is always 0 even in requests from old clients. | ||||
* | [SERVER] rpc: Support querying storage size + available space | Simon Rettberg | 2017-10-11 | 4 | -9/+27 |
| | |||||
* | [KERNEL] version check to support pointer to backing_dev_info in request_queue | Simon Rettberg | 2017-10-09 | 1 | -0/+6 |
| | | | | | This changed in 4.11, see https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=dc3b17cc8bf21307c7e076e7c778d5db756f7871 | ||||
* | [SERVER] Fix invalid pointer deref | Simon Rettberg | 2017-10-06 | 1 | -1/+1 |
| | |||||
* | [SERVER] rpc: Return a unique id generated at startup with each reply | Simon Rettberg | 2017-10-06 | 1 | -4/+19 |
| | |||||
* | [SERVER] Return client count when requesting stats but not client list | Simon Rettberg | 2017-10-06 | 2 | -4/+14 |
| | |||||
* | [SERVER] rpc: Fix bitmask calc for odd subnet sizes (not 8, 16, ...) | Simon Rettberg | 2017-10-04 | 1 | -1/+1 |
| | |||||
* | [SERVER] How about updating atime on connect? | Simon Rettberg | 2017-10-04 | 1 | -0/+3 |
| | |||||
* | [SERVER] Implement HTTP RPC that supports different queries and ACL | Simon Rettberg | 2017-10-01 | 5 | -45/+264 |
| | | | | | | | | | - ACL is defined in new file rpc.acl - Queries are still WIP, for now something like /query?q=stats&q=images /query?q=clients works, although the parsing is still ugly - Also supports HTTP keep-alive | ||||
* | [SERVER] Move includes from header to source unit | Simon Rettberg | 2017-10-01 | 2 | -2/+2 |
| | |||||
* | [SERVER] Helper function for loading line based space separated data from file | Simon Rettberg | 2017-10-01 | 3 | -34/+63 |
| | |||||
* | [*] constness for sock_sendAll params | Simon Rettberg | 2017-10-01 | 2 | -3/+3 |
| | |||||
* | [SERVER] Fix closing timeout reset, fix log messages | Simon Rettberg | 2017-09-08 | 2 | -1/+5 |
| | |||||
* | [SERVER] Refactor: Move client list to net.* and isolate | Simon Rettberg | 2017-09-07 | 8 | -247/+290 |
| | |||||
* | [SERVER] Implement closeUnusedFd config option | Simon Rettberg | 2017-09-02 | 4 | -0/+36 |
| | | | | | | | This will close the readFd of images that have no active clients after some idle period (1 hour currently). Prevents deleted images from taking up space until the server is shut down. | ||||
* | [*] Continue splitting #defines to clientconfig.h etc. | Simon Rettberg | 2017-09-02 | 10 | -35/+51 |
| | |||||
* | [SERVER] Split server-only defs from config.h to serverconfig.h | Simon Rettberg | 2017-08-26 | 8 | -32/+54 |
| | |||||
* | [SERVER] Change handling of nonworking images, check for size change | Simon Rettberg | 2017-08-24 | 1 | -77/+169 |
| | |||||
* | [SERVER] Fix status query on big endian | Simon Rettberg | 2017-08-21 | 1 | -4/+7 |
| | |||||
* | [SERVER] Make crclist handling endian-safe | Simon Rettberg | 2017-04-27 | 2 | -1/+9 |
| | |||||
* | [SERVER] Remove stray line | Simon Rettberg | 2017-04-27 | 1 | -1/+0 |
| | |||||
* | [SERVER] Fix image_load aswell... | Simon Rettberg | 2016-11-09 | 1 | -12/+12 |
| | |||||
* | [SERVER] Fix loading of rid-based images in vmdk legacy mode | Simon Rettberg | 2016-11-09 | 1 | -9/+8 |
| | |||||
* | [FUSE] Little more verbosity in some error msg | Simon Rettberg | 2016-11-09 | 1 | -1/+1 |
| | |||||
* | [KERNEL] PAGE_CACHE_SIZE -> PAGE_SIZE | Simon Rettberg | 2016-10-17 | 1 | -1/+1 |
| | | | | | | | These two have been the same all along and recently, the _CACHE variant has been removed (kernel 4.6). Switch to the non-CACHE version so build doesn't break on newer kernels. | ||||
* | SERVER_MAX_CLIENTS: 400 -> 4000 | konrad | 2016-07-19 | 1 | -1/+1 |
| | |||||
* | Added rudimentary benchmark tool. | Christian Klinger | 2016-07-15 | 6 | -0/+392 |
| | |||||
* | [SERVER] Bump debug lock tracking limits | Simon Rettberg | 2016-07-13 | 1 | -2/+2 |
| | |||||
* | [CLIENT] fix help output for closing a device | Jonathan Bauer | 2016-06-13 | 1 | -1/+1 |
| | |||||
* | [FUSE] Wait a little longer before switching via new approach | Simon Rettberg | 2016-03-23 | 1 | -8/+8 |
| | |||||
* | [FUSE] Add new load balancing mechanism | Simon Rettberg | 2016-03-23 | 1 | -12/+49 |
| | | | | | | | | | | | | | | | | | | | | | | | | | The new mechanism is supposed to complement the existing RTT based balancing. While the RTT-averaging approach is better suited to react to sudden drastic changes in server latency. The new approach doesn't directly consider RTT, but keeps track of how many consecutive times each server was the best server when measuring the RTTs. The higher that counter rises, the more likely it will become that the connection switches over to that server. Eg.: Server 1 measures 600µs each time Server 2 measures 599µs each time After a while, in case server 1 is currently used, the connection will eventually switch over to server 2. The RTT-based mechanism would not switch over in this case, since the threshold that prevents constant switching between servers is not reached. The new approach is meant to handle scenarios where the network is generally fast, but it would still be beneficial from a network topology point of view if the clients switch to the slightly faster server, assuming it is closer to the client and thus less network segments are burdened. | ||||
* | [SERVER] Fix formatting | Simon Rettberg | 2016-03-22 | 9 | -57/+58 |
| | |||||
* | [SERVER] (FreeBSD) Unlock send mutex on sendfile error | Simon Rettberg | 2016-03-22 | 1 | -0/+1 |
| | |||||
* | compile dnbd3-fuse on FreeBSD without warning :) | Sebastian | 2016-02-11 | 2 | -2/+4 |
| | |||||
* | remove kernel build breaking include | Sebastian | 2016-02-11 | 1 | -1/+0 |
| | |||||
* | running FreeBSD sendfile | Sebastian | 2016-02-11 | 1 | -13/+36 |
| | |||||
* | dnbd3server build success on freebsd :) | Sebastian | 2016-02-11 | 9 | -30/+77 |
| | |||||
* | cleanup commits | Sebastian | 2016-02-11 | 25 | -145/+47 |
| | |||||
* | [SHARED] signal.h -> fdsignal.h | Simon Rettberg | 2016-02-10 | 14 | -15/+15 |
| | |||||
* | dfskljgdslkfsdklgj | Simon Rettberg | 2016-02-05 | 1 | -3/+3 |
| | |||||
* | [SHARED] Fix linux detection on clang | Simon Rettberg | 2016-02-05 | 1 | -4/+4 |
| | |||||
* | [SHARED] Implement pipe() based signalling | Simon Rettberg | 2016-02-05 | 4 | -71/+261 |
| | |||||
* | First steps in make signals more abstract from the underlying mechanism; ↵ | Simon Rettberg | 2016-02-05 | 10 | -129/+136 |
| | | | | | | | | | | replace epoll with poll. We now don't assume that a signal equals a single fd (eventfd on Linux). The next step would be to create a version of signal.c that uses a pipe internally, so it can be used on other platforms, like *BSD. This is also the reason epoll was replaced with poll in uplink.c | ||||
* | [SERVER] BREAKING: Get rid of pseudo case-insensitivityv2.1 | Simon Rettberg | 2016-02-01 | 9 | -54/+41 |
| | | | | | | | | | | This was a wrong decision made long time ago, and it's broken in certain scenarios (eg. two servers serving from same NFS mount). Also it's of limited use anyways since it only supportes ASCII and would ignore umlauts, so blöd and BLÖD would still be considered two different images. So if you relied on this "feature" in any way, be careful when updating. | ||||
* | [SERVER] Fix a lot of (mostly harmless) data races | Simon Rettberg | 2016-01-11 | 13 | -94/+163 |
| | |||||
* | [KERNEL/CLIENT] Several minor tweaks and changes | Simon Rettberg | 2016-01-11 | 6 | -17/+20 |
| | | | | | | | | | | - Disable panic timeout that reports errors to block layer by default - Get rid of "is_server" used by old proxy mode, introduce "honor_server_alts" that just controls whether we use alts provided by the server - Allow switching servers when we only have 3 RTT measurements - If using alts provided by server, don't ignore those that come from other servers than the initial one |