| 9 Jan 2020 |
| 11:17:54 | |  eleitl joined the room. |
| 20:02:22 | ![@freenode__discord_417857235079790592[m]:matrix.org](./avatar/freenode__discord_417857235079790592%5Bm%5D) freenode__discord_417857235079790592[m] | Perhaps you could try |
| 20:02:50 | freenode__discord_417857235079790592[m] | What causes the scaling wall is that the pinset is stored as a DAG itself and requires walking it |
| 20:03:04 | freenode__discord_417857235079790592[m] | When you have hundreds of thousands of objects the walking can take a while |
| 20:03:32 | freenode__discord_417857235079790592[m] | Once you start getting up to around 3.5mil objects there appears to be about a 50% reduction in pinning performance |
| 20:04:10 | freenode__discord_417857235079790592[m] | The only solution is an overhaul of the pinning system |
| 20:09:47 | freenode__discord_417857235079790592[m] | Probably with reference counting tbh. I wrote a reference counter for a custom node I wrote and it's lightning fast |
| 10 Jan 2020 |
| 08:05:58 | eleitl | Thank you, this means there's a chance of it being fixed eventually. I can't be the only person who wants to pin more than a Tbyte on a single machine. |
| 17:20:17 | |  PHYQNZ joined the room. |
| 21:58:45 | freenode__discord_417857235079790592[m] | the one short term work around is to use chunksizes 1MB in size, so you'll reduce the number of pinned objects 4x compared to using 256kb chunk size |
| 23:19:29 | |  niels joined the room. |
| 11 Jan 2020 |
| 09:47:47 | eleitl | The problem is that it's going to change the CID, and I'm using the CID as a long-term registry ID for an immutable dataset. |
| 09:54:07 | | ![@freenode__discord_607248758790422538[m]:matrix.org](./avatar/freenode__discord_607248758790422538%5Bm%5D) freenode__discord_607248758790422538[m] joined the room. |
| 09:54:10 | freenode__discord_607248758790422538[m] | Can somebody help me with this: https://discuss.ipfs.io/t/ipfs-cluster-fails-to-post-to-ipfs-node/6977
? |
| 10:15:40 | freenode__discord_607248758790422538[m] | never mind guys..got it working 🙂 |
| 15:29:52 | eleitl | Am I running into any interoperability issues with the rest of the IPFS grid including cluster members if I'm picking 1 Mbyte for chunk size? |
| 13 Jan 2020 |
| 16:42:20 |  hsanjuan | eleitl: I think you more or less got answers to your questions. But if something is pending, would you mind asking on discuss.ipfs.io? Really hard to keep track of things here once I leave for a couple of days |
| 14 Jan 2020 |
| 15:00:33 | eleitl | Thanks, will do that in future. Also, will use GitHub to raise issues. |
| 15 Jan 2020 |
| 17:06:26 | |  @appservice-irc:matrix.orgchanged room power levels. |
| 17:06:39 | | @appservice-irc:matrix.orgchanged room power levels. |
| 16 Jan 2020 |
| 02:56:11 | | @appservice-irc:matrix.orgchanged room power levels. |
| 02:59:05 | | @appservice-irc:matrix.orgchanged room power levels. |
| 04:44:11 | | @appservice-irc:matrix.orgchanged room power levels. |
| 04:50:52 | | @appservice-irc:matrix.orgchanged room power levels. |
| 04:52:50 | | @appservice-irc:matrix.orgchanged room power levels. |
| 05:42:21 | | @appservice-irc:matrix.orgchanged room power levels. |
| 06:23:34 | | @appservice-irc:matrix.orgchanged room power levels. |
| 06:49:09 | | @appservice-irc:matrix.orgchanged room power levels. |
| 18 Jan 2020 |
| 18:52:35 | freenode__discord_607248758790422538[m] | does anybody know where is the IPFS Cluster config file stored in windows systems? |
| 23:52:43 | |  ferdinand joined the room. |
